The accuracy, reliability, and validity of the Medicare Peer Review Organization (PRO) review process have all been questioned.Evidence concerning the PRO program's effect on cost and quality of care remains lacking. The Health Care Financing Administration has thus committed itself to reform, and the Uniform Clinical Data Set (UCDS) has been proposed as the national database for Medicare's quality review program. The UCDS is an automated, computerized data set designed to standardize the evaluation of quality. It should allow an objective, consistent, and efficient process for peer review, based on explicit decision criteria and on aggregated information about patterns of care and quality. But is this truly so? We review the existing evidence on the UCDS and compare it with the current PRO reviews of quality, concluding that although the UCDS can potentially improve the accuracy and the reliability of data abstraction and the validity of reviews, this remains to be shown. Preliminary data on the UCDS suggest that work is needed before it can meet appropriate expectations for a national database for quality assessments. We also propose a model for reviews of quality in which we show that reviews of care done in the context of internal quality improvement programs will differ in goals and intensity from reviews of care done at the national level. We suggest that the UCDS has a unique, but limited role—that of national surveillance of practice patterns. Detailed assessments of quality are more appropriately done at local or institutional levels.