Forty hypertensive patients, twenty medicated and twenty umnedicated, were randomly distributed into four different groups, two experimental and two control groups: G1 (Unmedicated relaxation), G2 (Unmedicated control), G3 (Medicated relaxation) and G4 (Medicated control). The subjects included in the experimental groups were exposed to a muscle relaxation program which lasted eight weeks. The control groups were exposed to a weekly control of the blood pressure (BP) and heart rate (HR). Outcome measures included BP and HR in the clinic, measures of cardiovascular responsiveness (to mental arithmetic stress tests). The experimental groups G1 and G3 were superior to the control groups G2 and G4 in reducing BP in the clinic during post-treatment and follow-up. No significant differences were found between experimental groups. The differences between the experimental unmedicated group G1 and the control unmedicated group G3 was significant in reducing HR. In the same way, subjects belonging to the relaxation groups G1 and G3 showed a greater reduction in systolic and diastolic BP cardiovascular reactivity in relation to the control groups G2 and G4. The differences between the experimental unmedicated group G1 and the control unmedicated group G2 were significant in reducing HR during the mental stress. The results are relevant for the non-pharmacological treatment of essential hypertension.