1. Sixteen patients (11 male, five female), median age 41 years, with essential hypertension insufficiently controlled by hydrochlorothiazide (75 mg/day; diastolic blood pressure greater than or equal to 100 mmHg), were studied. 2. Plasma renin concentration [renin], plasma angiotensin II concentration ([ANG II]), plasma volume and exchangeable sodium (NaE) were determined, and a saralasin infusion (5.4 nmol min-1 kg-1) was carried out while the patients were on thiazide alone and, in 14 cases, 3 months after addition of a beta-adrenoreceptor blocker (propranolol, six, metoprolol, six, and atenolol, two patients). 3. On thiazide alone, saralasin caused a significant decrease in mean arterial blood pressure in 12 out of 16 patients. The saralasin response was closely related to pre-saralasin plasma [ANG II] (r = 0.73, P less than 0.01). Plasma [renin] and [ANG II] were higher than normal in the group as a whole. 4. After addition of a beta-adrenoreceptor blocker systolic and diastolic blood pressure decreased from 164/109 mmHg to 136/94 mmHg. Plasma [renin] and [ANG II] decreased by 40 and 58% respectively. At this point, saralasin caused no significant change in mean arterial pressure. No close correlation was found between plasma [renin] or [ANG II] or saralasin response on thiazide treatment and changes in blood pressure during subsequent thiazide/beta-adrenoreceptor-blocker treatment. Plasma volume and NaE did not change significantly. 5. In patients with thiazide-induced stimulation of the renin-angiotensin system, addition of a beta-adrenoreceptor blocker leads to suppression of the system and, at the same time, ANG II-dependence of blood pressure disappears. This contributes to the antihypertensive effect of beta-adrenoreceptor blockers in this particular situation.