The purpose of this study was to compare aortic blood pressure (AOR) to peripheral measurements by the Riva-Rocci/Korotkov (RRK) and Finapres continuous finger pressure (FIN) methods during dynamic and static exercise. A tip manometer was introduced in the ascending aorta after coronary angiography in 7 cardiac patients with good exercise capability. Static exercise was of moderate intensity and led to an increase of average diastolic and systolic AOR of 20 and 18 mmHg, respectively. The corresponding RKK values were 20 and 30 mmHg and the FIN values were 16 and 14 mmHg, respectively. In maximal cycle ergometry the discrepancies were larger, especially in the 4 subjects who reached 80% or more of predicted maximal work load. Diastolic and systolic increases in AOR in these 4 subjects were 12 and 38 mmHg, respectively. The RRK values were 17 and 76 mmHg. Increases in FIN values of 17 and 74 mmHg for diastolic and systolic measurements, respectively, were found. The peripheral FIN and RRK measurements give a systolic increase that is twice as large as that for AOR. It is concluded that RRK and FIN greatly overestimate the load to the cardiovascular system in dynamic exercise. When the cardiovascular load is estimated by the rate-pressure product, RRK produces an increase of 197%, FIN of 181%, while AOR gives an increase of only 133%. This suggests that the present criteria for blood pressure in exercise testing should be critically examined.