A criteria-based quality assurance process for hysterectomy was instituted at a large teaching hospital. After this process was initiated, the overall frequency of hysterectomy decreased by 24%, p < 0.001. Significant reductions were seen in hysterectomy rates for the following indications: chronic pelvic pain (77%, p < 0.0001), recurrent uterine bleeding (46%, p < 0.001), preinvasive disease of the uterus (55%, p < 0.005), and severe infection (70%, p < 0.025). Adenomyosis was the single indication for which an increase in hysterectomy rate was observed. This increase, however, was completely reversed during the last 2 years of the study. This quality assurance process also resulted in a significant increase in the histologic verification rate (i.e., 82% vs 93%, p < 0.001). These observations suggest that using such a criteria-based process can reduce the number of hysterectomies performed and improve the accuracy of the preoperative diagnosis. © 1990.