The therapeutic efficacy of long acting diltiazem 300 mg od (Tildiem LA) was compared with sustained release nifedipine 20 mg bd and bendrofluazide 5 mg od in a multicentre study with 230 patients diagnosed with mild to moderate essential hypertension, with 77, 77 and 76 randomized to the diltiazem, nifedipine SR and bendrofluazide groups respectively. Patients were entered into this randomised, single (investigator) blind, parallel-group multicentre study if the systolic and diastolic blood pressures were > or = 145 mm Hg and/or 95 mm Hg respectively at the admission visit. Twenty-one general practitioners and two hospital physicians monitored patients at baseline and at four and eight weeks of continuous dosing. After eight weeks of therapy, clinically acceptable control of blood pressure was seen in all groups: reductions were 19.2/13.5 mm Hg, 20.4/14 mm Hg and 18.5/10.8 mm Hg for the Tildiem, nifedipine and bendrofluazide groups respectively. Significant differences were shown between bendrofluazide and the other two groups on diastolic pressures (p = 0.01). The non-significant trend was for systolic pressures to mirror these effects. Significantly higher withdrawals caused by adverse events were seen with nifedipine. These were as follows: 14 patients receiving nifedipine (18%), 5 patients receiving diltiazem (6%) and 4 patients receiving bendrofluazide (5%). The difference in this withdrawal rate between treatments was statistically significant (p = 0.01). Post hoc tests revealed that both diltiazem and bendrofluazide had statistically significant lower withdrawals for adverse events than the nifedipine group (p = 0.047). Nifedipine was associated with a marginal increase in standing apex pulse rate and only diltiazem LA significantly maintained serum potassium levels. These results indicate that diltiazem 300 mg is an effective antihypertensive agent and is equivalent in efficacy to nifedipine SR 20 mg and both are superior to bendrofluazide. Nifedipine SR was however the worst tolerated and had the highest withdrawal rate (p = 0.013).