The preliminary results of an interdisciplinary active surveillance adverse drug reaction (ADR) reporting program conducted from April 1992 to December 1993 in the family medicine ward of National Taiwan University Hospital are presented. During this period, every admitted patient was screened for any possible drug-related problems; suspected ADRs were evaluated and documented. The USA Food and Drug Administration's definition of ADR was used, and Naranjo's scale was used to estimate the probability of drug-induced events. A total of 41 ADRs were identified in 38 patients. Among these 41 cases, 18 ADRs occurred during hospitalization, and 23 were the reason for hospitalization. The incidence of ADRs rated as probable or highly probable during hospital stay was 2.0%, or 2.7% if possible reactions were included (based upon 666 hospital admissions to the unit during the study period). The occurrence of admissions due to ADRs rated as highly probable or probable was 2.7%, or 3.5% if possible reactions were included. Twenty-one ADRs (51.2%) were type A reactions and 20 (48.8%) were type B. Thirty-three ADRs (80.5%) were classified as serious or moderate. There was no mortality. Further studies are warranted to determine the incidence of ADRs in different patient populations in Taiwan.