In a double blind, crossover study 6 h infusions of adrenaline (15 ng/kg/min; 1 ng = 5.458 pmol), noradrenaline (30 ng/kg/min; 1 ng = 5.911 pmol), and a 5% dextrose solution (5.4 ml/h), were given to ten healthy volunteers in random order 2 weeks apart. By means of intra-arterial ambulatory monitoring the haemodynamic effects were followed for 18 h after the infusions were stopped. Adrenaline, but not noradrenaline, caused a delayed and protracted pressor effect. Over the total postinfusion period systolic and diastolic arterial pressure were 6 (SEM 2)% and 7 (2)%, respectively, higher than after dextrose infusion (ANOVA, p less than 0.001). Thus, "stress" levels of adrenaline (230 pg/ml) for 6 h cause a delayed and protracted pressor effect. These findings are strong support for the adrenaline-hypertension hypothesis in man.