Mechanisms inducing a low cardiac output (CO) state in IgE-mediated (cytotropic) anaphylactic shock in anesthetized Macaca irus monkeys were studied. 7 monkeys were sensitized by 2 i.v. injections of a human reaginic serum containing a high concentration of IgE antibodies against dog albumin. Anaphylactic shock was elicited 48 h after the last sensitization dose, by an i.v. injection of dog albumin. The severe anaphylactic shock which developed was characterized by an initial phase consisting of increased CO (+16%, mean value, 1 min after challenge), pulmonary hypertension and systemic vasodilatation followed by a phase consisting of decreased CO (-67%), a fall in mean arterial pressure from 113 to 45 mmHg, decreases in left and right arterial pressures (-5.3 and -3.2 mmHg, respectively) and increases in pulmonary vascular (+364%) and total peripheral (+30%) resistances. These changes were recorded 5 min after challenge and the values then remained essentially unaltered during the rest of the 30-min observation period. Pulmonary vascular resistance was only increased by 140% at the end of that period. Myocardial blood flow was maintained during shock at the expense of flow to other organs. However, initially there was a redistribution of blood flow within the left ventricular myocardium, resulting in a relative decrease in subendocardial flow. This finding was not related either to occasional S-T changes in the electrocardiograms or to the level of decreased CO. The oxygen supply to the myocardium was reduced in shock but the reduction was always smaller than the corresponding decrease in heart work. Two additional monkeys sensitized with an IgE fraction from the human serum showed a smaller amount of specific IgG in their serum prior to challenge than the other monkeys. The response to challenge was milder, but resembled the initial vasodilatory reaction in monkeys sensitized with serum. These data on cytotropic anaphylaxis in the monkey show that the main cause of decreased CO and thus of the shock state is a decreased venous return, primarily due to peripheral blood pooling and, to a smaller extent, extravasation of plasma. No appreciable involvement of the heart in the induction of shock was detected.