The objectives of this study were to compare the effects of three different Pilates regimes on chronic, mild low back pain symptoms and to determine whether the efficiency of load transfer through the pelvis is improved by those exercises. A between subjects equivalent group experimental design was used. The independent variable was the type of exercise training (three groups) and the two-dependent variables were low back pain symptoms and load transfer through the pelvis. The outcome measures of the first-dependent variable were a comparison between modified Oswestry Disability Questionnaires (one of the standard pain instruments) completed pre- and post-program and frequency, intensity and duration of low back pain. The outcome measure of the second-dependent variable, efficiency of load transfer through the pelvis was the Stork test (one-legged standing test) in weight bearing. Although all groups experienced statistically significant reductions in frequency, intensity and duration of low back pain across the weeks of exercising, there were no significant differences between the groups relative to each other.