The evaluation of adverse drug reactions (ADR) in clinical practice is difficult and imprecise. Establishing a causal relationship may not be possible, and data on incidence cannot be obtained because the number of patients treated is not known. This article describes the ADR reported during the clinical trial program of cinoxacin, a synthetic antibacterial drug used to treat urinary tract infections. Results from 2,801 patients who received cinoxacin showed that 5 per cent reported ADR that were probably or definitely drug induced, and 10 per cent reported ADR in which the relationship was uncertain. There was no relationship between number of reports and patient's age, drug dose, or duration of treatment. Adverse drug reactions affecting the gastrointestinal system were reported by 5.5 per cent of the patients, those involving the central nervous system by 4.3 per cent, and hypersensitivity reported by 2.4 per cent. In the comparative studies, patients treated with cinoxacin reported fewer ADR than those treated with nalidixic acid, furadantin, amoxicillin, or trimethoprim-sulfamethoxazole. Although problems in the assessment and evaluation of ADR still exist, it is hoped that the results from the formal trial program will be representative of those seen in clinical practice.