Previous studies on the relationship between blood viscosity and athletic performance have involved the measurement of viscosity at 37[degrees]C. The aim of this study was to assess the effect of exercise-induced increases in core temperature on the viscosity of blood as this would more closely represent thein vivo situation. Eleven highly trained international level cyclists undertook two 30 minute time trials on a cycle ergometer. Core temperature was monitored during the trials using a rectal probe. Whole blood viscosity was measured using separate aliquots of the same venous sample in a Wells Brookfield cone/plate viscometer at both 37[degrees]C (WBV37) and at the actual core temperature of the subjects at the time of sampling(WBVCT). The mean increase in core temperature during the trials was 1.7+/-0.7[degrees]C (p<0.001). There was a significant increase in WBV37 during the exercise bouts (p<0.05), however the magnitude of the increase in WBVCT was 32% smaller than that observed in WBV37 (p<0.05). There was a significant reduction in plasma volume(PV) (Dill and Costill, 1974) during the trials (p<0.001) however there was no significant correlation between the change in PV and the increase in WBV at either temperature. These data indicate that the measurement of WBV at 37[degrees]C overestimates any exercise-induced increases and the measurement of WBVCT is important as it is likely to provide a more accurate indication of the in vivo behaviour of the blood.