BACKGROUND & AIMS: Levator ani syndrome (LAS) might be treated using biofeedback to teach pelvic floor relaxation, electrogalvanic stimulation (EGS), or massage of levator muscles. We performed a prospective, randomized controlled trial to compare the effectiveness of these techniques and assess physiologic mechanisms for treatment. METHODS: Inclusion criteria were Rome II symptoms plus weekly pain. Patients were categorized as "highly likely" to have LAS if they reported tenderness with traction on the levator muscles or as "possible" LAS if they did not. All 157 patients received 9 sessions including psychologic counseling plus biofeedback, EGS, or massage. Outcomes were reassessed at 1, 3, 6, and 12 months. RESULTS: Among patients with "highly likely" LAS, adequate relief was reported by 87% for biofeedback, 45% for EGS, and 22% for massage. Pain days per month decreased from 14.7 at baseline to 3.3 after biofeedback, 8.9 after EGS, and 13.3 after massage. Pain intensity decreased from 6.8 (0-10 scale) at baseline to 1.8 after biofeedback, 4.7 after EGS, and 6.0 after massage. Improvements were maintained for 12 months. Patients with only a "possible" diagnosis of LAS did not benefit from any treatment. Biofeedback and EGS improved LAS by increasing the ability to relax pelvic floor muscles and evacuate a water-filled balloon and by reducing the urge and pain thresholds. CONCLUSIONS: Biofeedback is the most effective of these treatments, and EGS is somewhat effective. Only patients with tenderness on rectal examination benefit. The pathophysiology of LAS is similar to that of dyssynergic defecation.