OBJECTIVE To clarify the spectrum of the left ventricular responses in hypertension in man by means of coupled analysis of elastances. In the present study we analysed the 'functional' coupling of the left ventricle and the arterial system in terms of the arterial effective elastance (E(a)) divided by a value of end-systolic left ventricular elastance approximated by the end-systolic pressure-volume ratio (E(lv)). METHODS Twenty-five normotensive and 19 hypertensive males without heart failure underwent a haemodynamic and angiographic study. The hypertensives were divided into three subgroups: group 1 had normal ejection fraction, group 2 had ejection fraction > 70% and group 3 had ejection fraction 50-58%. RESULTS The ejection fraction was similar in hypertensives and controls and E(lv) was significantly increased in the hypertensives. E(a) was identical in the three hypertensive subgroups, which differed only for E(lv). Hypertensives with a normal ejection fraction (n = 8) had a normal E(a)/E(lv) ratio and end-systolic stress, and a significantly increased E(lv), related mainly to an increase in the left ventricular mass divided by the end-diastolic volume (m/VED) with normal systolic function of the left ventricular muscle. The significantly increased systolic pump function of group 2 (n = 5) seems to be related to a significant increase in both m/VED and left ventricular muscle contractility. Group 3 (n = 6) was more heterogeneous, some patients having insufficient hypertrophy and others impaired muscle function. CONCLUSIONS The left ventricle and the arterial system remain correctly coupled in hypertensives overall, but with marked heterogeneity of the systolic pump (and sometimes muscle) function and mainly of the geometry of the left ventricle. Regarding the relatively unequivocal changes in Ea, the differences in ejection fraction and in left ventricular-arterial coupling in hypertensives are related mainly to changes in the left ventricular systolic pump function.