A dissociation between the hypotensive activity and the central side effects has been demonstrated with rilmenidine (RIL) in previous studies. The object of this trial was to compare RIL with methyldopa (MD) during a 12 weeks' treatment (Day 0 to Day 84) and after stopping therapy (Day 84-Day 91). In particular, the trial was designed to evaluate the acceptability of the two agents. After 4 weeks of placebo, 76 men and 81 women (average age 55 years) were allocated according to a randomised double-blind protocol to RIL (N = 78) or MD (N = 79). The two groups were comparable with systolic/diastolic BP of 165/102 mmHg and 165/101 mmHg respectively. Monotherapy was administered to all the patients from Day 0 to Day 56 in order to compare RIL (1 mg every morning or 1 mg twice daily if necessary) and MD (250 mg morning and evening or 500 mg twice daily if necessary). Hydrochlorothiazide (25 mg daily) was associated from Day 56 to Day 84 if the supine diastolic BP was greater than or equal to 90 mmHg. A placebo was substituted from Day 84 to Day 91 to study the effect of withdrawing therapy. RESULTS at Day 56 under monotherapy, the average supine systolic/diastolic BP had decreased significantly with RIL and MD to 140/110 mmHg and 140/100 mmHg respectively. An equivalent number of patients had normalised their blood pressures in both groups (systolic/diastolic BP less than or equal to 160/90 mmHg): 50% with RIL and 42% with MD (NS).(ABSTRACT TRUNCATED AT 250 WORDS)