Endothelin-1 was infused into the non-dominant brachial artery in two male subjects. We then monitored intra-arterial mean blood pressure, right atrial pressure, the heart rate and forearm blood flow (by plethysmography). Endothelin-1 at a dose of 5 x 1O-14 to 5 x 1O-9 mol, infused over 5 min periods, elicited no major changes in mean arterial pressure, heart rate and right atrial pressure. We observed an initial increase in forearm blood flow, followed by dose-dependent decreases of 25, 34 and 42% at 5 x 10-11 to 5 x 10-9 mol. A higher dose of endothelin-1, 5 x 10-8 mol, given to only one of the subjects, elicited sweating and vomiting. In this subject, mean arterial pressure, right atrial pressure and the heart rate did not change, while forearm blood flow increased transiently. A deep muscular pain developed in the forearm receiving the endothelin-1 infusion after 30 min (maximum 2h, duration 10 h), and this pain was intensified by touch and muscle contractions. The force of muscle contractions in the forearm was markedly reduced and a visible oedema developed. In order to investigate the mechanisms of oedema formation, endothelin-1 (10-10 to 5 x 10-8 mol/l) was given intra-arterially in a rat hindquarter preparation which was perfused at a constant flow rate. In the rat, endothelin-1 increased both pre-and postcapillary resistance, leading to an increase in capillary hydrostatic pressure and a marked net transcapillary fluid transfer from the perfusate to tissue. There was no sign of increased vascular permeability. These results indicate that the development of oedema in humans may be due to an increase in capillary pressure. The similarity to sarafotoxin toxicity suggests that endothelin peptides should be tested in humans with great caution until further data on their physiological or pharmacological properties are available.