Three hundred and thirty-three consecutive patients in a medical ward were evaluated in a high-intensity monitoring scheme for drug events as a cause of hospitalization. Taking into consideration only 'definite' and 'probable' drug events, we found 36 cases (10.8%) of all admissions to be drug-related hospitalizations (DRH). Of these, 8.1% were adverse drug reactions and 2.7% were therapeutic failures due to ineffective dosage. In 8 cases (2.4%) the drug event could definitely have been avoided, and a further 13 cases (3.9%) were considered to have been potentially avoidable if appropriate measures had been taken by the health service. In 19 cases (53%) the referring physician was unaware of the drug-related problem. Those patients admitted because of a drug event were taking significantly more drugs than other individuals. The avoidable drug events pointed to the primary health care physicians as the appropriate targets for preventive measures in terms of intensified drug education. The study demonstrated that a reliable estimate of the DRH rate requires active data collection by a qualified health service worker in close collaboration with the patient's family doctor in cases of suspected DRH.