In a double-blind double-dummy multicenter study, patients with mild to moderate essential hypertension were randomized to receive either nifedipine (n = 416, 47.6% women) or lisinopril (n = 412, 50% women), and side effects were registered by specific questioning, by spontaneous reports, and by use of visual analog scales. Cough was spontaneously reported to occur in 8.5% with lisinopril compared to 3.1% with nifedipine. Women treated with lisinopril reported cough spontaneously three times more often than men, 12.6% v 4.4%, whereas no differences between the sexes were observed during the placebo period or during nifedipine treatment. Similar gender differences were observed during specific questioning. Furthermore, nonsmokers reported an increase in cough more often than did smokers.