The purpose of this study was to investigate the immediate and delayed effects of plasma donation and blood donation on responses in exhaustive, severe-intensity exercise. Nineteen young men and women performed exhaustive cycle ergometer tests at ∼3.3 W·kg(-1) before and then 2 h, 2 days, and 7 days after withdrawal of either 8-10 mL·kg(-1) (∼700 mL) of plasma (n = 10) or 1 unit (450 mL) of whole blood (n = 9). Time to exhaustion was significantly (p < 0.05) decreased after the removal of plasma (-11% after 2 h) and after the removal of blood (-19% after 2 h and -7% after 2 days). Maximal oxygen uptake (.VO(2max)) was not affected by plasma donation, but .VO(2max) was reduced following blood withdrawal (-15% after 2 h, -10% after 2 days, and -7% after 7 days) presumably because of effects on blood volume, total haemoglobin content, and haemoglobin concentration. The kinetics of the oxygen uptake (.VO2) response was not affected by either intervention. Two measures of anaerobic capacity, postexercise blood lactate concentration, and maximal accumulated oxygen deficit were reduced (-14%, -15%, respectively) 2 h after plasma donation, but neither was affected by blood donation. Removal of plasma and removal of blood have different effects on blood constituency, on the .VO2 response, and on performance. Plasma donation appears to affect exercise performance because of reduced anaerobic capacity, whereas blood donation affects performance because of lowered .VO(2max).