OBJECTIVE: To compare clinical outcomes after randomization to hysterectomy versus medical treatment in patients with chronic abnormal uterine bleeding refractory to medroxyprogesterone acetate. METHODS: We randomly assigned 63 premenopausal women with abnormal uterine bleeding refractory to cyclic medroxyprogesterone acetate treatment to receive either a hysterectomy or expanded medical treatment. Within each randomized group, the specific treatment approach was determined by patient and provider preference. The primary analysis compared changes in clinical outcomes at 6 and 24 months by using an intention-to-treat approach. Secondary as-treated analyses after adjustment for baseline covariates compared participants randomly assigned to medical treatment who continued the medical approach with those who crossed over to hysterectomy. RESULTS: The intention-to-treat analyses at 6 months revealed greater symptom improvement in the hysterectomy group than in the medicine group for pelvic pain (P < .01), urinary urgency (P = .03), incomplete bladder emptying (P = .03), breast pain (P = .02), and cessation of vaginal bleeding (87% versus 11%, P < .001). Seventeen of 32 women assigned to medicine (53%) eventually crossed over and received a hysterectomy, and by 24 months the statistically significant differences by intention-to-treat were limited to greater improvement in hot flushes (P < .01) and cessation of vaginal bleeding (P < .01). Within-group analyses at year 2 showed statistically significant improvements from baseline on most symptoms for women who had a hysterectomy, whether through randomization or crossover. Women remaining on medical treatments had statistically significant improvements in pelvic pain, pelvic/bladder pressure, and stress incontinence. In a nonrandomized comparison with women who remained on medical treatments through year 2, those crossing over to hysterectomy experienced greater improvements in bleeding (P < .01), pelvic pain (P < .01), low back pain (P = .02), breast pain (P = .01), urinary frequency (P = .01), and urgency (P = .02). However, they also experienced more days off from work or usual activities (P < .01) and more days spent in bed (P < .01) than those who remained on medicine. CONCLUSION: For patients with abnormal uterine bleeding refractory to medroxyprogesterone acetate, hysterectomy is superior to expanded efforts with oral medications for alleviating clinical symptoms but may lead to more days of restricted activity. © 2004 by The American College of Obstetricians and Gynecologists.