OBJECTIVES: To describe changes in bladder function and voiding frequency associated with behavioral and drug treatment for urge incontinence and to examine whether these variables mediate the positive effects of treatment on the frequency of incontinence, DESIGN: Randomized, double-blinded, placebo-controlled, clinical trial. Eligible patients were stratified according to type of incontinence (urge only vs mixed urge and stress) and severity of incontinence (frequency of accidents as documented in bladder diary). SETTING: University-based outpatient geriatric medicine clinic. PARTICIPANTS: One hundred five ambulatory, nondemented, community-dwelling women; mean age 67.0 (range 55-91); 98% white, 2% African American. INTERVENTION: Four sessions (over 8 weeks) of biofeedback-assisted behavioral training, drug treatment with individually-titrated oxybutynin chloride, or a placebo control condition. MEASUREMENTS: Two-channel cystometry was performed to determine threshold volumes for first desire to void (FDV), strong desire to void (SDV), bladder capacity, and the volume at which detrusor instability (DI) or leakage occurred, before randomization and after completion of treatment. Two-week bladder diaries were used before and after treatment to document episodes of incontinence and voiding frequency. RESULTS: Bladder capacity increased by 68.9 mL in the oxybutynin group (P < .001) and 17.3 mL in the behavior group and decreased 6.0 mL in the control group. SDV increased 69.9 mL in the oxybutynin group (P < .001), 40.5 in the behavior group (P < .05), and 7.8 mL in the control group. FDV increased by 44.4 mL in the oxybutynin group (P < .001), 18.8 mL in the behavior group, and 8.9 mL in the control group. One of seven patients who presented with DI (12.0%) resolved in the behavior group, seven of eight (87.5%) resolved in the oxybutynin group, and seven of 12 (58.3%) resolved in the control group. These differences were not statistically significant. Voiding frequency was significantly reduced after treatment in both the behavior and the oxybutynin group. Behavioral training resulted in a mean 82.3% reduction in frequency of accidents, oxybutynin (final doses 2,5 mg daily to 5 mg three times a day) resulted in a mean 78.3% reduction, and the control condition resulted in a mean 51.5% reduction (P = .002). Although oxybutynin and behavioral treatment were both effective, and oxybutynin increased SDV and bladder capacity, the structural equation modeling did not demonstrate that the clinical improvement was mediated through the effects of these treatments on urodynamic or voiding frequency measures. CONCLUSIONS: Studies using more-complex urodynamics and studies with larger sample sizes are needed to better characterize changes in bladder function and explore other urodynamic changes that may accompany treatment. In addition, other factors, both physiological and behavioral, need to be explored as mechanisms by which conservative therapies improve urge incontinence.