The aims of this study were to examine familiarization and reliability associated with a 40-m maximal shuttle run test (40-m MST), and to compare performance measures from the test with those of a typical unidirectional multiple sprint running test (UMSRT). 12 men and 4 women completed four trials of the 40-m MST (8 × 40-m; 20 s rest periods) followed by one trial of a UMSRT (12 × 30-m; repeated every 35 s); with seven days between trials. All trials were conducted indoors and performance times were recorded via twin-beam photocells. Significant between-trial differences in mean 40-m MST times were indicative of learning effects between trials 1 and 2. Test-retest reliability across the remaining trials as determined by coefficient of variation (CV) and intraclass correlation coefficient (ICC) revealed: a) very good reliability for measures of fastest and mean shuttle time (CV = 1.1 - 1.3%; ICC = 0.91 - 0.92); b) good reliability for measures of blood lactate (CV = 10.1 - 23.9%; ICC = 0.74 - 0.82) and ratings of perceived exertion (CV = 5.3 - 7.6%; ICC = 0.79 - 0.84); and c) poor reliability for measures of fatigue (CV = 38.7%; ICC = 0.59). Comparisons between performance indices of the 40-m MST and the UMSRT revealed significant correlations between all measures, except pre-test blood lactate concentration (r = 0. 47). Whilst the 40-m MST does not appear to provide more information than can be gleaned from a typical UMSRT, following the completion of a familiarization trial, the 40-m MST provides an alternative and, except for fatigue measures, reliable means of evaluating repeated sprint ability.

• Tests of multiple sprint performance are a popular means of evaluating repeated sprint ability.

• Multiple sprint tests incorporating changes of direction may be more ecologically valid than unidirectional protocols.

• The 40-m maximal shuttle run test is a reliable way of evaluating repeated sprint ability following the completion of one familiarization trial.

• The 40-m maximal shuttle run test shows no clear advantage over a standard unidirectional multiple sprint test.