1. Reference values for systolic blood pressure during exercise are provided for 88 healthy adolescents (12-22 years of age) of both sexes. Data were related to oxygen uptake, heart rate, blood lactate concentration, rate of perceived exertion, age, sex, body size and physical fitness. 2. The same variables were measured in 55 adolescents of both sexes with insulin-dependent diabetes mellitus of about 12 years duration and were analysed with respect to the healthy control group, to degree of metabolic control and to late diabetic complications. 3. In healthy adolescents the pressure response was not related to sex or age. When compared with control subjects diabetic patients had a higher diastolic blood pressure at rest and a more marked blood pressure increase, 23 versus 19 mmHg W-1 kg-1 body weight, during exercise with no sex difference. The blood pressure rise was not related to metabolic control, glomerular hyperfiltration or physical fitness. 4. Prolonged exercise tests were no more informative regarding the blood pressure response to exercise than the stepwise increased load test. Analysing the blood pressure increase versus relative work load (W/kg body weight) during exercise reveals blood pressure differences otherwise not noted. A diabetic patient with blood pressure above the 97.5% confidence limit during exercise seems to have a higher risk of developing incipient nephropathy 5 years later.