Two hundred ninety-four consecutive admissions to a geriatric department were evaluated for drug events as cause of admission. The drug events considered were adverse drug reactions (ADRs) and dose-related therapeutic failures. In 39 cases (13.3%), a definite, probable or possible drug event was a dominant or contributing cause of the admission (11.2% ADRs and 2.0% dose-related therapeutic failures). Five of these cases (1.7%) were judged to be due to errors in prescription and a further seven (2.4%) were found to have been avoidable by efforts exceeding the obligatory. There were no statistically significant differences between drug-related and non-drug related admissions in terms of age, sex, number of drugs taken at the admission or duration of hospitalization. ADRs in the geriatric patients are difficult to recognize and may be interpreted as senile loss of function. Sixty-seven percent of the patients took drugs with a sedative action, 68% took drugs with a hypotensive action, 67% took drugs with a potential for confusional states, 15% took drugs with a potential for depression, and 64% took drugs with a potential for constipation. A reduction of the number of drug-related hospital admissions by means of a large-scale intervention would be a difficult task for several reasons: no particular class of drugs caused the drug events, no particular mechanism dominated, no particular group of doctors were responsible for the drug events, and only a part of the drug events were judged to be avoidable.