Compared the effectiveness of using a review team and the usual self-reporting method in detecting different types of medication errors in a state psychiatric hospital. Rates of prescription, transcription, administration, and dispensing errors were determined, and the risk of harm from each error was rated as high, moderate, or low. A review team was assigned to retrospectively review 31 patient records for prescription, transcription, and administration errors for a total of 1,448 patient-days. Dispensing errors, which can only be determined concurrently, were reported for an equivalent number of patient-days. The error rate was compared with the rate determined by self-reports from all nursing and medical staff. In the 31 charts reviewed and the dispensing events concurrently reviewed, the team detected a total of 2,194 medication errors, whereas a total of nine errors were self-reported for the same patient group. Administration errors accounted for more than half of the total, followed by transcription errors, prescription errors, and dispensing errors. 19% of errors were rated as having a low risk of harm, 23% as having a moderate risk, and 58% as having a high risk. Use of a review team should be considered as a method for detecting and reporting medication errors.