The purpose of this study was to evaluate a computerized information system, the Professional Activity Study-Medical Audit Program (PAS-MAP), when used by the medical staff of a hospital to conduct medical care evaluation studies. PAS-MAP was compared to a manual system for collecting data not contained on the face sheets of medical records. The results indicated that, compared to the manual system, PAS-MAP: was less costly if more than 41 per cent of hospitalized patients were included in medical care evaluation studies; was as timely as the manual system for data it could provide but provided fewer clinical data elements than physicians requested; and was less protective against human error. Three decision makers assigned weights indicating the relative importance of these results. The weights were combined in an additive model to arrive at a score for each system. Based on these scores, the manual system was recommended for implementation.