BACKGROUND Short-acting calcium antagonists may increase coronary artery morbidity, mortality, and non-cardiovascular complications in hypertensive patients. We assessed whether this association was also true for long-acting calcium antagonists. METHODS We did a case-control study, which was nested within a systematic hypertension control programme for a prospective cohort of 4350 people (first seen 1981-94). Cases (n = 189) were hypertensive patients who had had a first cardiovascular event, including all cardiovascular deaths and hospitalisations, between 1989 and 1995. Controls (n = 189) were individually matched to cases for sex, ethnic origin, age, type of previous antihypertensive treatment, year of entry into the study, and length of follow-up. We collected data on any prescribed drug regimen that was being taken on the event data for cases and on the same date for matched controls. Calcium antagonists were classified by duration of action. RESULTS Compared with those on beta-blocker monotherapy, patients on long-acting calcium antagonists (n = 136) had no increased risk of a cardiovascular event (adjusted odds ratio 0.76 [95% CI 0.41-1.43]), whereas patients on short-acting calcium antagonists (n = 27) were at significantly greater risk (adjusted odds ratio 3.88 [1.15-13.11], p = 0.029). Among 38 matched pairs who were both on calcium antagonists, the adjusted risk ratio for short-acting versus long-acting was 8.56 (1.88-38.97), p < 0.01. CONCLUSIONS Long-acting and short-acting calcium antagonists differ in cardiovascular outcomes. Consistent with earlier findings, the use of short-acting calcium antagonists was associated with increased risk of a cardiovascular event. This finding highlights the need to complete on-going clinical trials so that the relative cardiovascular impact of various antihypertensive agents can be assessed.