It is a general impression that the blood pressure (BP) response during monotherapy in hypertensive subjects is highly variable. As decreased insulin sensitivity is a frequent finding in hypertensive patients, the following study was performed to evaluate if the degree of insulin sensitivity could predict the BP response to different types of anti-hypertensive treatments. Insulin sensitivity was evaluated by the hyperinsulinaemic euglycaemic clamp technique before initiation of treatment with beta-adrenergic blockers (n = 181), thiazide diuretics (n = 60), ACE inhibitors (n = 73), non-dihydropyridine calcium antagonists (n = 38), dihydropyridine calcium antagonists (n = 26) or alpha-1 antagonists (n = 39) over periods of 3-6 months in hypertensive patients. The proportion of poor responders, defined as a reduction in the diastolic blood pressure (DBP) of < 3 mm Hg ranged between 8% and 30% in the different groups despite similar pretreatment DBPs (100-102 mm Hg). A decreased pretreatment insulin sensitivity was related to a poor DBP treatment response in the thiazide-treated group only (r = -0.33, P < 0.05). In this group also obesity, as evaluated by body mass index (BMI), was associated with a poor BP response (r = 0.28, P < 0.05), while obesity was a predictor of a favourable reduction in DBP in the group treated with non-dihydropyridine calcium antagonists (r = -0.34, P < 0.05). These associations were still significant when pretreatment DBP was taken into account in multiple regression analysis. Neither age nor sex were found to be significant predictors of BP response in any of the treatment groups.(ABSTRACT TRUNCATED AT 250 WORDS)