Epidemiological aspects of attacks of bronchial asthma related to drugs are prospectively studied in inpatients of three teaching hospitals in the Comprehensive Hospital Drug Monitoring (CHDM)-programme. Results are based on 34,840 individual patients (among 48,005 consecutive admissions) in the years 1974-1993. Between 1974 and 1993, every patient admitted to any of the three medical clinics in the CHDM programme was monitored for any suspicion of an adverse drug reaction (ADR); every drug exposure period during hospital stay was registered. Nineteen patients (0.05% of the 34,840 individual patients) had at least one attack of bronchial obstruction during hospitalisation, considered as probable or definite ADR. The frequency related to exposure periods in response to penicillins is 0.014%, to non-steroidal anti-inflammatories (NSAIDs) 0.0145, to acetyl salicylic acid (ASA) 0.018%, to paracetamol 0.008% and to beta-adrenoceptor blockers 0.26%. Of the 12 patients reacting to a drug with an allergic or idiosyncrasy/intolerance type of bronchial obstruction, 7 had a history of bronchial asthma (extrinsic or intrinsic), and 3 had the diagnosis chronic obstructive pulmonary disease (COPD). A history of bronchial asthma or COPD is confirmed to be a risk factor for this particular ADR. Of the seven patients with a bronchial obstruction to beta-adrenoceptor blockers, five were diagnosed with COPD, while two had neither COPD nor bronchial asthma. The relative risk for this pharmacological reaction in COPD patients was 96 (95% confidence interval 45-208) compared with non-COPD patients in the group of 3244 exposed to beta-adrenoceptor blockers.