Anterior cruciate ligament (ACL) injury is one of the most common knee injuries in athletic populations (Chia et al., 2022), with an incidence of approximately 1.5 per 10, 000 athlete-exposures (Montalvo et al., 2019). For athletes aiming to return to preinjury level of sports, ACL reconstruction (ACLR) is commonly considered as the first-line treatment option (Feucht et al., 2016; Paudel et al., 2023). However, current evidence indicates that approximately 20% of patients are unable to resume their preinjury level of sports and 45% cannot return to competitive sports after ACLR (Connors et al., 2025). Additionally, the risk of reinjury during the return to sports (RTS) period remains substantial, with rates as high as 12-19% (Liu et al., 2026). These suboptimal outcomes may be influenced by multiple factors; however, the marked heterogeneity in existing RTS evaluation frameworks highlights the need to improve RTS criteria as a potential strategy to enhance postoperative recovery and reduce reinjury risk.

Determining an athlete’s readiness for RTS is a complex, multifactorial clinical decision (Turk et al., 2023). Existing RTS criteria commonly include postoperative time, patient-reported outcome measures, isometric and isokinetic assessments of quadriceps and hamstring strength, hop tests, and evaluations of psychological readiness (Turk et al., 2023; Wright et al., 2025). Although no universal consensus has been established, functional performance assessments are considered both valid and reliable in guiding RTS decisions (Aguero et al., 2024; Glattke et al., 2023). A systematic review of post-ACLR assessment strategies demonstrated that functional performance testing is the most frequently implemented criterion (98.4% implementation rate), with hop tests being the most widely used functional testing modalities (Roe et al., 2022).

As the most commonly used assessment, hop testing has been recognized as a reliable evaluation tool to estimate functional recovery following knee injuries (Andrade et al., 2020; Berg et al., 2022). It can provide both quantitative measures (e.g., hopping distance and symmetry index) and qualitative movement descriptors (e.g., motor control patterns and biomechanics) to guide rehabilitation progression (Davies et al., 2020; Kumar et al., 2025) and inform RTS decision-making (Aguero et al., 2024). Nevertheless, evidence regarding the associations between hop test performance and successful RTS as well as a secondary ACL injury remains limited and inconclusive (Ashigbi et al., 2020; Cronström et al., 2023). These inconsistent findings may be attributed to the variability in hop test protocols, with systematic reviews identifying more than 15 different testing approaches reported in the literature (Burgi et al., 2019; Roe et al., 2022). Accordingly, selecting a hop test protocol that can effectively differentiate patients who can successfully and safely RTS after ACLR is crucial for enhancing the screening capacity of clinical RTS criteria.

Therefore, the primary purpose of this systematic review was to evaluate whether hop test performance is associated with successful RTS following ACLR. The secondary purpose was to determine whether hop test performance is effective at identifying reinjury occurrences after ACLR.

This systematic review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Page et al., 2021) and was registered prospectively with PROSPERO (CRD42025638911) on 29 January 2025.

Studies were included if they met the following criteria: (1) recruited participants who had undergone primary and unilateral ACLR surgery, with no restrictions on age, surgical techniques or graft type; (2) applied at least one hop test (quantitative measure) to assess functional performance. Hop testing was defined as a functional performance task in which the individual executes a countermovement to generate propulsive force, achieves a flight phase, and completes the action with a stabilized landing on one or both limbs; (3) reported outcomes on whether participants experienced RTS or sustained ACL reinjuries with no restrictions on follow-up duration. RTS in this review was defined as sport participation at the same level as before the injury, while reinjury was defined as either ipsilateral graft rupture or contralateral ACL injury, with no restrictions on assessment approaches; (4) examined the association between hop test measures and RTS or reinjuries.

Studies were excluded for the following reasons: (1) inclusion of participants with concomitant injuries other than meniscal, cartilage and medial collateral ligament (MCL) injuries, as these injuries are prevalent in individuals with ACL injuries (Christino et al., 2019; Niknam et al., 2025; Rotterud et al., 2011); (2) publication type being conference proceedings, case reports, clinical commentaries or review studies.

The electronic databases PubMed, MEDLINE, SPORTDiscus with Full Text, and Web of Science were systematically searched from inception to 28 February 2026. The search strategy combined MeSH terms and keywords including the following terms and their variations: (“anterior cruciate ligament” OR ACL) AND (reconstruction OR surgery OR operation) AND (“return to play” OR “return to sport” OR “return to activity” OR “return to performance” OR RTS OR RTP) AND (hop OR jump OR land) (Supplementary Table 1). After removing duplicates, titles and abstracts were screened against the inclusion criteria by two independent reviewers (Y.Y.L. and Y.X.S.), followed by independent full-text assessments using the same criteria. Any discrepancies were resolved through discussion and consultation with the corresponding author (J.H.Q.).

Two authors (Y.Y.L. and Y.X.S.) independently evaluated the methodological quality of the included studies using a modified version of the Newcastle-Ottawa scale (West et al., 2023), which comprised ten items across four domains: participant selection, definition of exposure, comparability of cohorts and outcome assessment. Each item was rated as high or low risk of bias.

In addition, a modified Grade of Recommendations Assessments, Development and Evaluation (GRADE) framework (Montgomery et al., 2025) was used to assess the certainty of evidence for each outcome. This tool rated an initial evidence certainty as ‘high’, downgraded by 5 items (study limitations, inconsistency, imprecision, indirectness, and reporting bias) and upgraded by 2 items (pooled effect size and exposure-response gradient) (Supplementary Table 2). Reporting bias was further examined using funnel plots and Egger’s test when at least 10 studies were available for a given outcome in the meta-analysis.

Data were extracted independently by two authors (Y.Y.L. and Y.X.S.). Extracted variables from each individual study included participant characteristics (age, sex, graft type, etc.), hop test types, hop test measures (absolute distance, limb symmetry index [LSI], etc.), RTS clinical criteria, rates and timing of RTS or reinjury, determination methods of RTS and reinjury, follow-up duration, reported associations between hop testing and RTS and/or reinjury. The LSI represents the performance of the injured limb relative to the uninjured limb, which is calculated as;

When necessary, corresponding authors of the included studies were contacted to obtain additional information.

Meta-analyses were conducted when at least 3 sets of data for a given outcome were available for pooling (Review Manager, Version 5.3. Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2014) and when the odds ratios (ORs) with standard errors (SEs) could be calculated. The ORs were used to estimate the strength of associations between hop test performance and the probability of successful RTS or reinjury. If ORs were not reported or could not be derived from outcome frequencies, the standard mean difference in hop test performance between participants with and without successful RTS/reinjury was calculated and converted to an OR (Higgins et al., 2019). A random effect model was used to account for expected heterogeneity in graft types, outcome measures and follow-up durations.

Statistical heterogeneity was evaluated for each outcome using Cochran’s Q-test and quantified with the I² statistic, which represents the proportion of total variability due to heterogeneity rather than chance (Higgins et al., 2019). Unplanned subgroup analyses and meta-regression were performed to explore potential sources of heterogeneity, such as the presence versus absence of concomitant meniscal injuries.

When data were insufficient for meta-analysis, results were synthesized through Synthesis Without Meta-analysis (SWiM) (Campbell et al., 2020).

The initial search identified 7870 articles. After removing duplicates and sequentially screening titles, abstracts and full-texts, 43 studies were included in this systematic review (Figure 1), involving a total of 17, 350 participants with ACLR, with 38.87% being female. Of the included studies, 36 were cohort studies, five were cross-sectional studies, and one was case-control study. Thirty-five (81.4%) individual studies reported that participants engaged in at least a recreational level of sports before the ACL injury (Table 1). The majority of the participants (90.17%) received hamstring tendon (HT) autografts, while other graft types included quadriceps tendon (QT, 0.38%), bone-patellar tendon-bone (BPTB, 8.46%), allograft (0.88%) and synthetic graft (0.11%). In addition, 23.63% of participants sustained concomitant meniscal injuries and 7.79% had concomitant MCL injuries among studies reporting concomitant injuries (Table 1).

Across the included studies, 17 distinct hop tests were documented. The three most frequently utilized were the single-leg hop for distance (SLHD) (38 studies, 86.4%), triple hop for distance (THD) (19 studies, 43.2%), and triple crossover hop for distance (TCHD) (13 studies, 30.2%). A total of 7 measures of hop test performance were reported in the relevant studies, including LSI, side-to-side difference, absolute distance, hopping numbers, hopping time, hop measures normalized to height and to body mass, with LSI being the most widely applied in 39 studies (90.7%) (Table 2).

Hop performance data were collected between 6 and 24 months postoperatively, whereas follow-up for RTS/ reinjury spanned 7 months to 11 years post-ACLR. Beyond hop tests, the assessment modalities in RTS criteria across the individual studies demonstrated homogeneity. Specifically, 38 studies (88.3%) consistently assessed patient-reported outcomes and 33 (76.7%) tested maximal knee-extensor or -flexor strength (Supplementary Table 3).

Twenty-four studies (Bodkin et al., 2022; Faleide et al., 2021; Ithurburn et al., 2019; Jang et al., 2014; Kim et al., 2022; Kitaguchi et al., 2020; Legnani et al., 2024; Moksnes and Risberg, 2009; Müller et al., 2015; Nawasreh et al., 2018; Niederer et al., 2025; Niederer et al., 2024; Novaretti et al., 2018; Ohji et al., 2024; Ohji et al., 2021a; Ohji et al., 2021b; Senorski et al., 2017; Toole et al., 2017; Ueda et al., 2022; Webster and Feller, 2017; Webster and Feller, 2018; Webster and Feller, 2020; Webster et al., 2019; Welling et al., 2020) investigated the association between hop testing and RTS, and 21 (Bodkin et al., 2022; Cristiani et al., 2022; Cristiani et al., 2025; Fältström et al., 2023; Figueroa Poblete et al., 2025; Girdwood et al., 2024; Grindem et al., 2016; Ithurburn et al., 2019; Kew et al., 2022; King et al., 2021; Kyritsis et al., 2016; Marigi et al., 2023; Marty-Diloy et al., 2025; Nyland et al., 2025; Paterno et al., 2025; Paterno et al., 2022; Piussi et al., 2023; Simonsson et al., 2024; Van Melick et al., 2022; Webster and Feller, 2019; Zsidai et al., 2026) examined reinjury outcomes. RTS status was determined by self-report in 12 studies (50.0%) and by quantitative score in 12 studies (50.0%). Reinjury was by self-report in 7 studies (33.3%) and via MRI or medical records in 8 studies (38.1%) (Supplementary Table 3).

Among all the included studies, only 4 (9.3%) conducted power analysis. Sixteen studies (37.2%) had a dropout rate > 15% during follow-up (mean dropout rate across all included studies was 10.03%), and 24 (55.8%) did not control for potential confounders (e.g., age, sex, preinjury sports level) in the statistical analysis (Supplementary Table 4).

A total of 24 studies (Bodkin et al., 2022; Faleide et al., 2021; Ithurburn et al., 2019; Jang et al., 2014; Kim et al., 2022; Kitaguchi et al., 2020; Legnani et al., 2024; Moksnes and Risberg, 2009; Müller et al., 2015; Nawasreh et al., 2018; Niederer et al., 2025; Niederer et al., 2024; Novaretti et al., 2018; Ohji et al., 2024; Ohji et al., 2021a; Ohji et al., 2021b; Senorski et al., 2017; Toole et al., 2017; Ueda et al., 2022; Webster and Feller, 2017; Webster and Feller, 2018; Webster and Feller, 2020; Webster et al., 2019; Welling et al., 2020) investigated the association between hop test performance and the success of RTS following ACLR, reporting 13 different types of hop tests. In the relevant studies, the performance of hop tests was typically assessed within 6 to 24 months postoperatively, with subsequent RTS evaluation occurring from 7 to 34.4 months post-ACLR (Table 2).

Nineteen studies (Bodkin et al., 2022; Ithurburn et al., 2019; Jang et al., 2014; Kim et al., 2022; Kitaguchi et al., 2020; Moksnes and Risberg, 2009; Müller et al., 2015; Nawasreh et al., 2018; Niederer et al., 2024; Ohji et al., 2024; Ohji et al., 2021a; Ohji et al., 2021b; Senorski et al., 2017; Toole et al., 2017; Ueda et al., 2022; Webster and Feller, 2017; Webster and Feller, 2018; Webster and Feller, 2020; Welling et al., 2020) (79.2%) employed the SLHD to evaluate participants’ functional performance. Different outcome measures were used to quantify the performance across studies, including LSI (n = 16), absolute distance (n = 2), side-to-side difference (n = 1) and distance normalized to height or body mass (n = 4) (Table 2). A meta-analysis of 15 studies (Bodkin et al., 2022; Kitaguchi et al., 2020; Moksnes and Risberg, 2009; Müller et al., 2015; Nawasreh et al., 2018; Niederer et al., 2024; Ohji et al., 2024; Ohji et al., 2021a; Ohji et al., 2021b; Toole et al., 2017; Ueda et al., 2022; Webster and Feller, 2017; Webster and Feller, 2018; Webster and Feller, 2020; Welling et al., 2020) (5315 participants) demonstrated that achieving an LSI ≥ 90% in SLHD was significantly associated with successful RTS (OR = 1.20, 95% CI: 1.09 to 1.33, I² = 78%) (Figure 2), whereas one study (Ithurburn et al., 2019) with qualitative analysis reported no association (Table 3). Overall, low certainty of evidence suggests that achieving an LSI ≥ 90% on SLHD is significantly associated with the success of RTS following ACLR (Supplementary Table 5).

Regarding other quantitative measures of SLHD performance, the certainty of evidence remains inconsistent or limited. For the absolute distance of SLHD, one (Welling et al., 2020) demonstrated that greater distance on both involved and uninvolved limbs significantly correlated with RTS, while the other (Jang et al., 2014) showed no association. For height-normalized SLHD distance, two (Ithurburn et al., 2019; Ohji et al., 2021b) of three studies revealed that better performance on the involved limb was associated with successful RTS, while the third (Bodkin et al., 2022) found no association; only one (Ithurburn et al., 2019) examined the performance of uninvolved limb and found no association. No association results were also reported by one study (Senorski et al., 2017) which examined body-mass-normalized distance on both involved and uninvolved limbs. Only one study (Kim et al., 2022) measured SLHD performance by side-to-side difference and found a significant association (Table 3, Supplementary Table 5).

Seven studies (Bodkin et al., 2022; Ithurburn et al., 2019; Moksnes and Risberg, 2009; Müller et al., 2015; Nawasreh et al., 2018; Toole et al., 2017; Welling et al., 2020) (29.2%) examined the association between THD performance and RTS, using different outcome measures: LSI (n = 7), absolute distance (n = 1) and distance normalized to height (n = 2) (Table 2). A meta-analysis of 6 studies (Bodkin et al., 2022; Moksnes and Risberg, 2009; Müller et al., 2015; Nawasreh et al., 2018; Toole et al., 2017; Welling et al., 2020) (520 participants) demonstrated no association between achieving an LSI ≥ 90% and successful RTS (OR = 1.07, 95% CI: 0.99 to 1.16, I² = 71%) (Figure 2), which was consistent with the findings of an additional study (Ithurburn et al., 2019) included in qualitative synthesis (Table 3). Therefore, low certainty of evidence suggests that achieving an LSI ≥ 90% on THD had no association with successful RTS following ACLR (Supplementary Table 5).

Among studies that did not measure LSI, one study (Welling et al., 2020) reported that a greater absolute distance on both involved and uninvolved limbs in THD was associated with a higher probability of RTS. Two additional studies (Bodkin et al., 2022; Ithurburn et al., 2019) yielded inconsistent findings regarding the association between height-normalized THD distance and successful RTS on involved and uninvolved limbs. Specifically, Ithurburn et al. (2019) found a significant association on both involved and uninvolved limbs, whereas Bodkin et al. (2022) reported no association on involved limb (Table 3). Overall, the available evidence regarding absolute/normalized distance of THD and RTS following ACLR was limited and inconsistent (Supplementary Table 5).

Seven (Ithurburn et al., 2019; Moksnes and Risberg, 2009; Müller et al., 2015; Nawasreh et al., 2018; Toole et al., 2017; Webster and Feller, 2017; Webster and Feller, 2020) of the 24 studies (29.2%) evaluated the association between TCHD performance and RTS, using LSI (n = 7) or distance normalized to height (n = 1) (Table 2). A meta-analysis of 6 studies (Moksnes and Risberg, 2009; Müller et al., 2015; Nawasreh et al., 2018; Toole et al., 2017; Webster and Feller, 2017; Webster and Feller, 2020) (3081 participants) demonstrated that achieving an LSI ≥ 90% in TCHD was significantly associated with successful RTS (OR = 1.18, 95% CI: 1.02 to 1.36, I² = 69%) (Figure 2), while no association was reported by one additional study (Ithurburn et al., 2019) in the qualitative synthesis (Table 3). The above findings reached a low certainty of evidence suggests that achieving an LSI ≥ 90% in TCHD was significantly associated with successful RTS (Supplementary Table 5). Furthermore, a single study (Ithurburn et al., 2019) indicated that higher height-normalized TCHD distance on both involved and uninvolved limbs was associated with RTS (limited certainty of evidence) (Table 3, Supplement Table 5).

Four studies (Bodkin et al., 2022; Moksnes and Risberg, 2009; Nawasreh et al., 2018; Toole et al., 2017) (16.7%) examined the association between 6-mTH performance and RTS using the measures of LSI (n = 4) or hopping time (n = 1) (Table 2). A meta-analysis of 4 studies (Bodkin et al., 2022; Moksnes and Risberg, 2009; Nawasreh et al., 2018; Toole et al., 2017) (417 participants) indicated that achieving an LSI ≥ 90% in 6-mTH was not associated with successful RTS (OR = 1.02, 95% CI: 0.98 to 1.06, I² = 74%), which achieved a low certainty of evidence (Figure 2, Supplementary Table 5). A non-significant association was reported in a single study (Bodkin et al., 2022) that measured hopping time (limited certainty of evidence) (Table 3, Supplementary Table 5).

Eight studies (Legnani et al., 2024; Müller et al., 2015; Niederer et al., 2025; Niederer et al., 2024; Novaretti et al., 2018; Ohji et al., 2024; Senorski et al., 2017; Welling et al., 2020) evaluated the association between measures of eight individual hop tests and probability of RTS, including side hop, balance front hop, drop vertical jump, single-leg vertical jump (SLVJ) and square hop, etc. (Table 2).

For side hop test, one study (Welling et al., 2020) reported that a greater hopping number on both involved and uninvolved limbs was associated with RTS. However, no significant associations were found between body mass normalized hopping numbers and RTS (Senorski et al., 2017). In addition, low certainty of evidence from two studies suggested that the LSI in side hop was not associated with RTS (Legnani et al., 2024; Welling et al., 2020) (Table 3, Supplementary Table 5).

Regarding the LSI in SLVJ and single-leg drop vertical jump, limited evidence from one study (Ohji et al., 2024) found significant differences between RTS and non-RTS groups for both tests (Table 3, Supplementary Table 5).

Evidence for hop test batteries was inconsistent: two studies (Nawasreh et al., 2018; Webster et al., 2019) observed that meeting an LSI ≥ 90% in hop test batteries was associated with RTS, whereas one study (Faleide et al., 2021) found no association (Table 3). For the remaining hop tests, no significant associations between performance measures and successful RTS were observed, with the certainty of evidence ranging from limited to low (Table 3, Supplementary Table 5).

Twenty-one studies (Bodkin et al., 2022; Cristiani et al., 2022; Cristiani et al., 2025; Fältström et al., 2023; Figueroa Poblete et al., 2025; Girdwood et al., 2024; Grindem et al., 2016; Ithurburn et al., 2019; Kew et al., 2022; King et al., 2021; Kyritsis et al., 2016; Marigi et al., 2023; Marty-Diloy et al., 2025; Nyland et al., 2025; Paterno et al., 2025; Paterno et al., 2022; Piussi et al., 2023; Simonsson et al., 2024; Van Melick et al., 2022; Webster and Feller, 2019; Zsidai et al., 2026) evaluated the association between hop test performance and the occurrence of reinjury after ACLR, reporting a total of 12 different hop tests. The hop testing was normally conducted between 6 and 19 months post-ACLR, while reinjury follow-up periods ranged from 2 to 11 years after surgery (Table 2).

Seventeen studies (Bodkin et al., 2022; Cristiani et al., 2022; Cristiani et al., 2025; Fältström et al., 2023; Girdwood et al., 2024; Grindem et al., 2016; Ithurburn et al., 2019; Kew et al., 2022; King et al., 2021; Kyritsis et al., 2016; Marigi et al., 2023; Marty-Diloy et al., 2025; Paterno et al., 2025; Paterno et al., 2022; Simonsson et al., 2024; Webster and Feller, 2019; Zsidai et al., 2026) (81.0%) investigated the association between SLHD performance and reinjury, among which LSI (n = 16), absolute distance (n = 4) and height-normalized distance (n = 4) were used to quantify the performance (Table 2). A meta-analysis of 13 studies (Bodkin et al., 2022; Cristiani et al., 2022; Cristiani et al., 2025; Fältström et al., 2023; Grindem et al., 2016; Kew et al., 2022; King et al., 2021; Kyritsis et al., 2016; Marigi et al., 2023; Paterno et al., 2022; Simonsson et al., 2024; Webster and Feller, 2019; Zsidai et al., 2026) (9393 participants) found that achieving an LSI ≥ 90% in SLHD was not associated with odds of reinjury (OR = 1.02, 95% CI: 0.99 to 1.06, I² = 29%) (Figure 3), supported by two studies (Ithurburn et al., 2019; Paterno et al., 2025) excluded from this meta-analysis (Table 4). Overall, moderate certainty of evidence suggests that achieving an LSI ≥ 90% in SLHD was not associated with the risk of ACL reinjury following ACLR (Supplementary Table 5).

Furthermore, the results of another meta-analysis of 3 studies (Bodkin et al., 2022; Fältström et al., 2023; Marigi et al., 2023) involving 616 participants showed no significant association between height-normalized SLHD distance on the involved limb and reinjury (OR = 1.02, 95% CI: 0.97 to 1.06, I² = 68%). This finding was further supported by 2 additional studies (Girdwood et al., 2024; Ithurburn et al., 2019) included in the qualitative synthesis, providing low certainty of evidence for the absence of an association (Figure 3, Table 4). Similarly, five studies (Fältström et al., 2023; Ithurburn et al., 2019; King et al., 2021; Marigi et al., 2023; Marty-Diloy et al., 2025) consistently reported no association between SLHD performance and reinjury odds when measured by absolute or height-normalized distance (Table 4). Overall, this evidence was rated as low certainty (Supplementary Table 5).

Ten studies (Bodkin et al., 2022; Grindem et al., 2016; Ithurburn et al., 2019; Kew et al., 2022; Kyritsis et al., 2016; Marigi et al., 2023; Marty-Diloy et al., 2025; Nyland et al., 2025; Paterno et al., 2025; Paterno et al., 2022) (47.6%) investigated the association between THD performance and reinjury, using LSI (n = 9), absolute distance (n = 2) and height-normalized distance (n = 2) as outcome measures (Table 2). A meta-analysis of 7 studies (Bodkin et al., 2022; Grindem et al., 2016; Kew et al., 2022; Kyritsis et al., 2016; Marigi et al., 2023; Nyland et al., 2025; Paterno et al., 2022) (1178 participants) indicated that achieving an LSI ≥ 90% in THD was significantly associated with increased odds of reinjury (OR = 1.05, 95% CI: 1.01 to 1.09, I² = 0%) (Figure 4). However, two additional studies (Ithurburn et al., 2019; Paterno et al., 2025) that were not included in the meta-analysis reported no association (Table 4). Overall, the certainty of evidence for this association was rated as moderate (Supplementary Table 5).

Two studies (Marty-Diloy et al., 2025; Nyland et al., 2025) observed no association between greater absolute distance and reinjury odds on either involved or uninvolved limbs (very low certainty of evidence). Regarding height-normalized distance, two studies (Bodkin et al., 2022; Marigi et al., 2023) found no association with reinjury, whereas one study (Ithurburn et al., 2019) reported significant associations on both involved and uninvolved limbs (inconsistent certainty of evidence) (Table 4, Supplementary Table 5).

Seven studies (Grindem et al., 2016; Ithurburn et al., 2019; Kew et al., 2022; Kyritsis et al., 2016; Paterno et al., 2025; Paterno et al., 2022; Webster and Feller, 2019) (33.3%) investigated the association between TCHD performance and reinjury, assessing LSI (n = 7) or distance normalized to height (n = 1) (Table 2). A meta-analysis of 5 studies (Kew et al., 2022; Kyritsis et al., 2016; Paterno et al., 2022; Webster and Feller, 2019) (804 participants) demonstrated that there was no association between achieving an LSI ≥ 90% in TCHD and reinjury occurrence (OR = 1.00, 95% CI: 0.84 to 1.20, I² = 16%) (Figure 4). This finding was also supported by one study (Ithurburn et al., 2019) with qualitative synthesis. Overall, low certainty of evidence suggests that there was no significant association between LSI ≥ 90% in TCHD and ACL reinjury risk.

Two studies (Paterno et al., 2025; Webster and Feller, 2019) further investigated the predictive effects of TCHD for the occurrence of subsequent ipsilateral graft ruptures and contralateral ACL injuries and found no association with either outcome, which reached a low certainty of evidence. Only one study (Ithurburn et al., 2019) reported no association between height-normalized TCHD distance and reinjury odds on either involved or uninvolved limbs (limited certainty of evidence) (Table 4, Supplementary 5).

Five studies (Bodkin et al., 2022; Grindem et al., 2016; Kew et al., 2022; Paterno et al., 2025; Paterno et al., 2022) (23.8%) examined the association between 6-mTH performance and reinjury, using LSI (n = 5) and hopping time (n = 1) as outcome measures (Table 2). A meta-analysis of these 4 studies (Bodkin et al., 2022; Grindem et al., 2016; Kew et al., 2022; Paterno et al., 2022) (472 participants) found no association between achieving an LSI ≥ 90% in 6-mTH and reinjury (OR = 1.00, 95% CI: 0.90 to 1.11, I² = 20%) (Figure 4). This finding was consistent with one additional study (Paterno et al., 2025). Overall, the certainty of evidence was rated as low (Supplementary Table 5). Additionally, limited evidence from a single study (Bodkin et al., 2022) evaluating hopping time (in seconds) also reported no association (Table 4, Supplementary Table 5).

Nine studies (Fältström et al., 2023; Figueroa Poblete et al., 2025; Girdwood et al., 2024; King et al., 2021; Marigi et al., 2023; Marty-Diloy et al., 2025; Nyland et al., 2025; Simonsson et al., 2024; Zsidai et al., 2026) investigated the association between reinjury occurrence and 7 types of individual hop test performance, including side hop, 5-jump, single-leg drop jump, SLVJ, etc. (Table 2).

For the side hop, a meta-analysis of 3 studies (Fältström et al., 2023; Simonsson et al., 2024; Zsidai et al., 2026) (694 participants) found no significant association between achieving an LSI ≥ 90% and reinjury odds (OR = 1.00, 95% CI: 0.97 to 1.02, I² = 0%) (Figure 4), and the certainty of evidence was moderate. Limited-certainty evidence from one study (Fältström et al., 2023) reported significant differences in height-normalized hopping numbers between reinjury and non-reinjury groups.

For absolute hopping numbers on both involved and uninvolved limbs, findings were inconsistent: two studies (Girdwood et al., 2024; Marty-Diloy et al., 2025) found no significant differences, in contrast with one additional study (Fältström et al., 2023) (Table 4, Supplementary Table 5). Similarly, a meta-analysis of 3 studies (Marigi et al., 2023; Simonsson et al., 2024; Zsidai et al., 2026) involving 921 participants found no association between achieving an LSI ≥ 90% in SLVJ and reinjury odds (OR = 1.03, 95% CI: 0.94 to 1.12, I² = 71%) (Figure 4). The certainty of evidence was rated as low. In addition, a single study (Marigi et al., 2023) also reported no association between height-normalized jump length in SLVJ and reinjury (Table 4), providing limited certainty of evidence (Supplementary Table 5).

In addition, limited certainty of evidence from a single study (Fältström et al., 2023) suggested that both jump length and height-normalized length in the 5-jump test differed significantly between reinjury and non-reinjury groups (Table 4, Supplementary Table 5).

For hop test batteries, low certainty of evidence indicated no significant association with reinjury. Although one study (Van Melick et al., 2022) reported that achieving an LSI ≥ 90% in a hop battery was associated with decreased odds of reinjury, four additional studies (Figueroa Poblete et al., 2025; Paterno et al., 2022; Piussi et al., 2023;

Simonsson et al., 2024) reported no association despite using different combinations of hop tests. Similarly, no significant associations were found between the remaining hop test measures and reinjury risk, and the certainty of evidence ranged from limited to low (Table 4, Supplementary Table 5).

Subgroup analyses regarding graft type, concomitant meniscal injury, timing of hop test, measures of RTS and follow-up duration for the association between LSI performance of SLHD and RTS were conducted to address significant heterogeneity, which was not feasible for other hop tests due to inadequate data for stratification. A significant subgroup difference was observed according to the proportion of participants receiving HT autografts (I² = 75.1%, p = 0.05), with stronger association found in studies with > 90% HT autograft use (OR = 1.64, 95% CI: 1.13 to 2.39, I² = 82%) compared with those with ≤ 90% (OR = 1.11, 95% CI: 1.01 to 1.21, I² = 74%) (Figure 5). In addition, subgroup analysis found achieving an LSI ≥ 90% in SLHD was associated with increased likelihood of RTS in participants without meniscal injury (4412 participants; OR = 1.40, 95% CI: 1.15 to 1.70, I² = 83%), but not in those with meniscus injury (903 participants; OR = 1.14, 95% CI: 0.98 to 1.34, I² = 70%) (Figure 5). Additional subgroup analyses confirmed consistent significant associations between achieving an LSI ≥ 90% in SLHD and RTS probability across all other subgroups, regardless of timing of hop testing (≥ 1 year postoperatively vs. < 1 year postoperatively), measures of RTS (quantitative scores vs. patient-reported dichotomous outcomes) and follow-up duration (> 2 years vs. ≤ 2 years) (Figure 6, Supplementary Figure 2).

Furthermore, random-effects meta-regressions were performed to investigate the potential influence of study-level variables on the association between the LSI in SLHD and RTS. In univariable analyses, no study-level variables were significantly associated with effect size. In multivariable meta-regression, the model including meniscus status and follow-up duration was statistically significant (p = 0.006). Both meniscus status (β = -0.520, p = 0.003) and follow-up duration (β = -0.047, p = 0.002) were independently associated with effect size. However, across all models, the proportion of between-study variance explained by the moderators was negligible (R² = 0.00%), and considerable residual heterogeneity remained (residual I² range: 65.83% to 80.77%) (Supplementary Table 6).

Additionally, to determine whether meeting an LSI ≥ 90% of SLHD had different effects on the prediction of ipsilateral graft rupture and contralateral ACL injury, additional meta-analyses were conducted. Neither outcome showed a significant association. Specifically, for ipsilateral graft rupture, three studies (Cristiani et al., 2022; King et al., 2021; Webster and Feller, 2019) (n = 526) reported a pooled OR of 1.10 (95% CI: 0.97 to 1.25, I² = 0%); for contralateral ACL injury, three studies (Kew et al., 2022; Kyritsis et al., 2016; Webster and Feller, 2019) (n = 5787) reported a pooled OR of 1.34 (95% CI: 0.95 to 1.90, I² = 0%) (Figure 3).

Sensitivity analysis using the leave-one-out method revealed that the pooled effect estimates remained significantly associated between SLHD LSI and RTS (Supplementary Table 7). Although all I² values exceeded 70%, the findings indicate that the overall heterogeneity was not driven by any single study, thereby supporting the robustness of this association. A significant association was also found (9 studies, OR = 1.30, 95% CI: 1.05 to 1.62, I² = 77%) when dropping studies with a high risk of bias, strengthening this robustness (Supplementary Figure 1). However, the funnel plot with trim-and-fill imputation (Duval and Tweedie, 2000) suggested potential publication bias (imputed 6 studies; OR = 1.06, 95% CI: 0.95 to 1.18) (Supplementary Figure 4), which may have influenced the observed association between LSI of SLHD and RTS. This analysis was not performed for other outcomes because of insufficient data.

This review found that achieving an LSI ≥ 90% in SLHD and TCHD was significantly associated with a higher likelihood of return to preinjury level of sports; achieving an LSI ≥ 90% in THD was significantly associated with an increased risk of reinjury following ACLR. Associations between other hop test measures and RTS, as well as reinjury risk, remained nonsignificant, inconsistent or limited. Given the small effect size, the association between the LSI in THD and reinjury should be interpreted with appropriate caution.

This review found increased probability of RTS among individuals who achieved an LSI ≥ 90% in SLHD and TCHD (Figure 2), which aligns with previous reviews (Ardern et al., 2014; Gill et al., 2024b; Losciale et al., 2020; West et al., 2023). Although the effect sizes were modest, subgroup and sensitivity analyses demonstrated the robustness of the findings. However, these results should be interpreted cautiously due to the identified between-study heterogeneity and potential publication bias. Therefore, while SLHD and TCHD may provide clinically relevant information regarding RTS, they should be considered as components of a comprehensive, multidimensional RTS assessment rather than standalone criteria.

In contrast, achieving LSI ≥ 90% in THD was not significantly associated with RTS or 6-mTH (Figure 2), which differed from previous evidence (West et al., 2023). The discrepancy may be attributed to differences in study populations. West et al. (2023) recruited individuals with ACL injuries treated either operatively or non-operatively, whereas the present review focused exclusively on those treated operatively. Differences in recovery trajectories and rehabilitation strategies among these two populations (Grindem et al., 2014) may influence muscular endurance and fatigue resistance, thereby affecting performance on longer-duration hop tests (Komi, 2000; Kotsifaki et al., 2022a).

The association between SLHD symmetry and RTS appeared to be influenced by several clinical factors. Our subgroup analyses identified graft type as a potential effect modifier. Specifically, the association between SLHD LSI ≥ 90% and RTS was stronger in the subgroup with the use of HT autografts > 90% (OR = 1.64) compared with the ≤ 90% subgroup (OR = 1.11), with high between-subgroup heterogeneity (I² = 75.1%) (Figure 5). This finding may partly reflect differences in functional recovery after different graft choices, as patients receiving BPTB autografts have been reported to experience a higher risk of anterior knee pain and impaired quadriceps neuromuscular function, which may negatively influence psychological confidence and sports participation (Hui et al., 2011; Spindler et al., 2004). However, this finding should be interpreted as an effect of populations predominantly receiving different graft types rather than a direct effect of graft choice itself.

In addition, the subgroup analyses found that achieving an SLHD LSI ≥ 90% was significantly associated with RTS in individuals without meniscal injury, but not in those with meniscal injury (Figure 5). The meta-regression further supported this finding, showing a significant negative association between meniscal injury and RTS effect estimates (Supplementary Table 6). Given the role of the meniscus in load distribution, knee stability, and proprioception (Flandry and Hommel, 2011; Fox et al., 2012), concomitant meniscal injuries may impair neuromuscular control and functional recovery, reducing the ability of hop symmetry to discriminate RTS outcomes. Therefore, patients with combined ACLR and meniscal injury may require additional assessments beyond hop test performance, such as lower-limb muscle strength, movement quality, psychological readiness, and other established RTS criteria. Further investigation should provide more detailed reporting of meniscal treatment status to better evaluate its potential influence on hop performance and RTS outcome.

Furthermore, our meta-regression analysis also identified follow-up duration as a significant moderator (Supplementary Table 6), suggesting that the association between SLHD symmetry and RTS may diminish over time. This finding appeared to be strengthened by subgroup analyses, as the strength of the association was weaker in studies with longer follow-up duration (Figure 6). As neuromuscular control, muscle strength, and functional performance are largely restored within the first 2 years following ACLR (Buckthorpe et al., 2024), factors influencing long-term RTS decisions become more multifaceted and complex, such as fear of reinjury, restoration of sport-specific skills, and return to high-level competition (Nagelli and Hewett, 2017). However, these results should be interpreted cautiously due to the limited number of studies (5 studies), high dropout rate (20.02%) during follow-up, and the between-study heterogeneity that the identified moderators did not explain (R² = 0%).

Our subgroup analyses and meta-regression did not identify potential covariates that explained the substantial between-study heterogeneity, with R² of 0% across all models and persistently high residual I² values (Supplementary Table 6). This suggests that additional unmeasured factors, such as patient characteristics (e.g., sports type and activity level), surgical factors (e.g., reconstruction technique and meniscal treatment), and rehabilitation adherence, may further contribute to the observed variability. Future high-quality studies should therefore incorporate objective RTS measures, adequately control for relevant covariates and confounders, and report absolute risk estimates alongside relative measures to better establish the clinical utility of hop testing.

Consequently, these findings support the inclusion of SLHD or TCHD tests as essential components of a multifaceted RTS assessment criterion, particularly for patients undergoing HT autograft reconstruction, without concomitant meniscal injuries during the first 2 years after ACLR. Due to substantial heterogeneity, potential publication bias and limited certainty of evidence, future longitudinal studies are needed to better clarify their roles in making RTS decisions.

In contrast to previous systematic reviews (Cronström et al., 2023; Losciale et al., 2020), our review identified a statistically significant association between achieving an LSI ≥ 90% in THD and an increased risk of reinjury (Figure 3). However, this finding should be interpreted cautiously, given the small effect size, limited number of studies, and inconsistency with previous evidence. The RTS criteria used in the included studies were generally homogeneous beyond hop testing (Supplementary Table 3), suggesting that differences in non-hop-test evaluations were unlikely to explain the observed association. Additionally, the homogeneity (I² = 0%, Cochran’s Q = 4.18, p = 0.65) should be interpreted cautiously, as it may be limited in statistical power, given the limited number of studies. Furthermore, very-low to moderate certainty and lack of significant associations between other hop test outcomes and reinjury indicate that hop test alone provides limited value for identifying reinjury risk after ACLR.

The limited association between hop tests and reinjury risk may be explained by the inability of traditional hop test metrics, such as LSI and distance, to capture underlying deficits in movement quality that are more related to ACL loading (Kotsifaki et al., 2020; Xergia et al., 2015). Non-contact ACL injury, accounting for the majority of ACL injuries (Della Villa et al., 2021; Lucarno et al., 2021), occurs as rapidly as 17 to 50 milliseconds after initial ground contact (Koga et al., 2010; Krosshaug et al., 2007) and aberrant biomechanical patterns, such as excessive knee valgus, greater asymmetry in knee extensor moments, and decreased knee flexion angles, may contribute to elevated ACL loading (Gill et al., 2024a; Peebles et al., 2022). However, traditional hop test outcomes, such as LSI or hopping distance, primarily quantify overall performance rather than movement quality during the landing phase (Kotsifaki et al., 2020). Moreover, LSI in hop tests may overestimate functional recovery because the contralateral limb also exhibited persistent neuromuscular deficits (Girdwood et al., 2024) and movement quality abnormalities may remain despite meeting symmetry thresholds (Kotsifaki et al., 2022a; Kotsifaki et al., 2022b). Therefore, hop-test symmetry may not provide sufficient sensitivity to identify biomechanical deficiencies relevant to reinjury prevention.

Another potential explanation for the limited association is that laboratory-based hop tests may not adequately represent the complexity of competitive sports. Laboratory- or clinical-based hop tests are typically performed in predictable, single-plane conditions and do not incorporate sport-specific challenges, such as opponent interaction (McLean et al., 2004), unplanned movement (King et al., 2018) and divided attention (Almonroeder et al., 2019). Moreover, psychological factors, including fear of reinjury and low self-efficacy, may influence movement quality (Zhou et al., 2023), readiness for RTS (Niederer et al., 2025; Paterno et al., 2025) and functional performance (Beischer et al., 2019), which may further be associated with reinjury occurrence.

In addition, the best passing cutoff values for hop tests remain uncertain. Although an LSI threshold of 90% is widely used (Barber-Westin and Noyes, 2011; Thomeé et al., 2011), persistent functional and biomechanical deficits have been observed in individuals who meet this criterion (Gill et al., 2024a; Kotsifaki et al., 2022a; Kotsifaki et al., 2022b). Further research using longitudinal designs and appropriate predictive modelling is needed to establish clinically meaningful thresholds. Furthermore, future studies should analyze ipsilateral graft rupture and contralateral ACL injury separately rather than treating them as a single reinjury outcome, as these outcomes may involve different mechanisms and not be equally predicted by hop-test performance (Cronström et al., 2023; Cronström et al., 2021).

Consequently, LSI or distance-related measures on hop tests may be unable to predict reinjury risk after ACLR. Future research may move beyond isolated hop performance and develop multidimensional risk assessment models incorporating environmental adaptability and movement quality to improve reinjury risk stratification and promote safer RTS decisions.

This review has several limitations. Firstly, the certainty of evidence, particularly for RTS outcomes, was rated low or very low under the GRADE framework, primarily due to the significant heterogeneity and potential reporting bias. High heterogeneity may be partly explained by graft type, but likely stems from unmeasured clinical or methodological factors. Reporting bias is largely based on the inadequately repeated investigation. Additionally, participant attrition during follow-up may have introduced potential attrition bias. Therefore, high-quality, large-scale prospective cohort studies are warranted to strengthen the evidence. In addition, our search strategy prioritized studies reporting RTS as a primary outcome. This focus may have excluded some reinjury-focused cohort studies not emphasizing RTS in titles or abstracts. However, comparison with prior systematic reviews showed our study selection was comparable in scope, indicating that the findings were not materially influenced by the search strategy. Finally, we were unable to determine whether surgical factors, such as graft type and concomitant meniscal procedures, modified the associations between hop test performance and RTS or reinjury because these variables were inconsistently reported and data were limited.

Low certainty of evidence suggests that achieving an LSI ≥ 90% in single-leg hop for distance and triple crossover hop for distance is associated with a higher likelihood of RTS following ACLR, particularly within the first two years following ACLR and among individuals reconstructed with hamstring autografts and without concomitant meniscal injuries. Moderate certainty of evidence suggests a modest association between achieving an LSI ≥ 90% in triple hop for distance and a higher likelihood of ACL reinjury. Although these associations were small in magnitude, the findings suggest that functional performance assessed by hop tests may reflect readiness for RTS but does not necessarily correspond to a lower risk of reinjury, highlighting the limitations of relying solely on hop performance when making RTS decisions. Therefore, clinicians may consider incorporating SLHD/TCHD LSI values as part of a broader, multifactorial RTS assessment framework. Incorporating movement quality, neuromuscular control, and contextual factors may improve reinjury risk stratification and enhance safer RTS decision-making.