A prospective study was carried out in 2 groups of hypertensive patients (482 inpatients followed during 2-3 wk; 120 outpatients followed for 18 mo.) to evaluate the prevalence of side effects (SE) induced by antihypertensive drugs in current use. Only SE of definite or probable type as operatively defined were taken into consideration. SE were collected in different ways for inpatients and outpatients. Among 120 outpatients, 30 patients were randomly selected for cross over of placebo and drugs. Definite or probable SE were recorded in 236 (48.9%) of 482 patients: the prevalence of SE was below 20% both in single and multiple drug therapy. A prevalence greater than 20% was recorded in the outpatients. The placebo treatment demonstrated that many SE are not related to drug administration. Drugs studied included clonidine, dihydrochlorothiazide, methyl dopa, propranolol, dihydralazide and amiloride.