Eight competitive cyclists [mean peak oxygen consumption, (VO2(peak)) = 65 ml x min(-1) x kg(-1)] undertook two 60-min cycle ergometer time trials at 32 degrees C and 60% relative humidity. The time trials were split into two 30-min phases: a fixed-workload phase and a variable-workload phase. Each trial was preceded by ingestion of either a glycerol solution [1 g x kg(-1) body mass (BM) in a diluted carbohydrate (CHO)-electrolyte drink] or a placebo of equal volume (the diluted CHO-electrolyte drink). The total fluid intake in each trial was 22 ml x kg(-1) BM. A repeated-measures, double blind, cross over design with respect to glycerol was employed. Glycerol ingestion expanded body water by approximately 600 ml over the placebo treatment. Glycerol treatment significantly increased performance by 5% compared with the placebo group, as assessed by total work in the variable-workload phase (P < 0.04). There were no significant differences in rectal temperature, sweat rate or cardiac frequency between trials. Data indicate that the glycerol-induced performance increase did not result from plasma volume expansion and subsequently lower core temperature or lower cardiac frequencies at a given power output as previously proposed. However, during the glycerol trial, subjects maintained a higher power output without increased perception of effort or thermal strain.