2275 hypertensive patients consulting their usual general practitioner were analyzed using a specifically designed questionnaire. Sixty-two percent of the subjects were already treated and 38% were untreated prior to inclusion in the study. The data, collected during the usual medical consultation, included several questions answered by the physician concerning the patient's psychological status as well as a series of behavioural scales. The aim of this study was to describe a small number of different and homogeneous psychological patterns rather than a unique profile of hypertensive out-patients. Four types were thus defined using a statistical clustering procedure and were designated as ‘everything is OK”, “hyper-reactive”, “depressed”, and “dynamic-competitive”, representing respectively 43%, 33%, 10% and 14% of the population studied. The first type consists of patients who do not express subjective complaints and claim to have no conflicts, but who have difficulty adjusting to environmental changes. The second type is characterized by emotionally unstable individuals who are easily overwhelmed by their difficulties. The third type corresponds to pessimistic subjects, still affected by previous painful life events. Finally, the fourth type, close to coronary-prone type A behaviour, is composed of subjects whose aggression is channelled into social competition. The distribution of hypertensives amongst these 4 types is the same whether previously treated or not. Under treatment throughout the 6 months’ follow-up, the evolution of mean systolic blood pressure (SUP) and of mean diastolic blood pressure (DBP) was similar in the 4 types, although the second type showed the highest SBP during this period. Second and third type subjects more frequently reported the classical symptoms of hypertension. First and third type subjects reported more side-effects with antihypertensive treatment. Each type's vulnerability to stress is discussed from a psychosomatic viewpoint of essential hypertension. This study provides evidence that several behavioural patterns can be described amongst hypertensive patients. Knowledge of these patterns may be very useful in clinical practice to manage the Patient-physician relationship, and to increase patient compliance with antihypertensive therapy.