The incidence and prevalence of cough related to enalapril was assessed by spontaneous reporting and a visual analogue scale during a 6 month random double-blind parallel-group study comparing enalapril with nifedipine. Cough was reported spontaneously by 6.2% of enalapril-treated patients, and by none on nifedipine (NS). No patient had to discontinue enalapril because of cough. After 24 weeks treatment increases in visual analogue scale scores for cough frequency greater than or equal to 8 mm were more common for enalapril than nifedipine (difference 21.5%, 95% CI 7.3-35.7%). Increased cough frequency by visual analogue scale was present throughout the study in women, but less consistently in men. High scores for cough were not related to the dose of enalapril. Cough with enalapril was not an important problem during the 6 months of treatment. However increased cough frequency could be detected by visual analogue scale, with a frequency consistent with that observed in open clinic-based studies of longer duration. These findings suggest that ACE inhibitor-induced cough may increase in severity over time, and that even a period of 6 months treatment is too short to evaluate this side-effect adequately.