This study examined the effectiveness of Ballates training (strengthening of the central core musculature by the inception of balance techniques) compared to more traditional exercise programs, such as step aerobics and walking, on balance in women aged 50- 75 years. Participants were randomly assigned to one of three supervised training groups (1 hour/day, 3 days/week, 13 weeks), Ballates (n = 12), step aerobics (n = 17), or walking (n =15). Balance was measured by four different methods (modified Clinical Test for the Sensory Interaction on Balance - mCTSIB; Unilateral Stance with Eyes Open - US-EO or Eyes Closed - US-EC; Tandem Walk - TW; Step Quick Turn - SQT) using the NeuroCom Balance Master. A 2-way (Group and Trial) repeated measures ANOVA and post-hoc Bonferroni Pair-wise Comparisons were used to evaluate changes in the dependent variables used to describe stability and balance (sway velocity, turn sway, speed, and turn time). Measures of static postural stability and dynamic balance were similar for the three groups prior to training. Following the different exercise interventions, sway velocity on firm and foam surfaces (mCTSIB) with eyes closed (p < 0.05) increased for the Ballates group while the other two exercise groups either maintained or decreased their sway velocity following the training, therefore suggesting that these two groups either maintained or improved their balance. There were significant improvements in speed during the TW test (p < 0.01), and turn time (p < 0.01) and sway (p < 0.05) during the SQT test for each of the three groups. In general, all three training programs improved dynamic balance, however, step aerobics and walking programs resulted in be better improvements in postural stability or static balance when compared to the Ballates program.