To evaluate a 10-week training program for improved lumbo-pelvic stability (LPS), and to investigate the impact of improved LPS on athletic performance.Randomized-controlled study.Testing was done at the University of British Columbia, Vancouver. Training was at colleges and universities in the immediate vicinity to Vancouver, Canada.Thirty female varsity volleyball and basketball players aged 18–23yrs were randomly assigned to treatment (T), pseudo-treatment (PT), or control (C).For LPS the participants lied supine and the position of the pelvis and lumbar spine was monitored using a stabilizer Pressure Biofeedback Unit™ (PBU) while load was progressively added by movements of the lower limbs. T-test, Sargent's, and Bass' tests assessed agility, leg power, and balance, respectively.Non-parametric Friedman, Wilcoxon and Mann–Whitney techniques detected LPS improvement in T (2.8±1.5) and PT (2.3±1.4). Repeated measures ANOVA detected improvement in the agility (8.8±0.7s) and leg power (32.3±4.5cm) of T, and in the static balance ability of all three groups. Regression using Spearman's rho revealed no significant correlations between the post-test scores for LPS and athletic performance, or between pre- to post-test changes in LPS was improved following training in the T and PT groups. While improvements in agility and leg power were limited to only the T group, there was no association between improvements in LPS and improvements in athletic performance.The PBU may be an important tool in identifying lumbo-pelvic instability, however, its use is not recommended in the evaluation of treatment efficacy.