The function of both right and left sides of the heart was studied during spontaneous attacks of angina pectoris at rest in 7 patients showing ST depression (type I) and 4 showing ST elevation (type II) during the attack. In none of the 44 type I attacks and 29 type II attacks which were recorded did circulatory changes; the latter were different in the two groups. Type I attacks showed: a) a brief fall in arterial pressure, accompanied by b) a rise of right atrial and pulmonary wedge pressures and c) a decrease of cardiac output, right and left stroke work, the mean rate of systolic ejection, and indirect left ventricular pre-ejection dP/dt. In the course of the attack a hypertensive phase followed, which was paralleled by an increase of heart rate, cardiac output, left and right stroke work, and mean systolic ejection rate, left dP/dt; right atrial pressure and wedge pressure remained raised. All of the circulatory functions started to revert towards the pre-attack levels coincident with the waning phase of the electrocardiographic alteration, the latter occurring either spontaneously or after nitroglycerin. Type II attacks for the entire duration of the electrocardiographic changes showed: a) a reduction of arterial pressure, cardiac output, right and left stroke work, mean systolic ejection rate, and left dP/dt, b) a rise of right atrial and wedge pressures, and c) quite small changes of heart rate. When the electrocardiogram started to revert to the pre-attack aspect, the cardiac function rapidly improved and, after a supernormal phase, returned to the basal levels in about 2 minutes. It is concluded: 1) that no circulatory factor interfering with the mechanical effort of the heart is responsible for eliciting spontaneous angina: 2) that in type I attacks right and left ventricular impairment occurs which recovers rapidly, possibly through a sympathetic compensation; 3) that in type II attachs dysfunction of both sides of the heart occurs and persists throughout the episode of electrocardiographic alteration; 4) that the dynamic impairment is probably more severe in type I than in type II angina.