Eight endurance-trained cyclists rode as far as possible in 1 h on a stationary cycle simulator in a moderate environment (20 degrees C, 60% relative humidity, 3 m.s-1 wind speed) while randomly receiving either no fluid (NF) or attempting to replace their approximate 1.71 sweat loss measured in a previous 1-h familiarisation performance ride at approximately 85% of peak oxygen uptake with artificially sweetened, coloured water (F). During F, the cyclists drank mean 1.49 (SEM 0.14) 1 of which mean 0.27 (SEM 0.08) 1 remained in the stomach at the end of exercise and mean 0.20 (SEM 0.05) 1 was urinated after the trial. Thus, only mean 1.02 (SEM 0.12) 1 of the ingested fluid was available to replace sweat losses during the 1-h performance ride. That fluid decreased the mean average heart rate from 166 (SEM 3) to 157 (SEM 5) beats.min-1 (P < 0.0001) and reduced the final mean serum [Na+] and osmolalities from 143 (SEM 0.6) to 139 (SEM 0.6) matom.l-1 (P < 0.005) and from 294 (SEM 1.7) to 290 (SEM 1.9) mosmol.l-1 (P = 0.05), respectively. Fluid ingestion did not significantly attenuate rises in plasma anti-diuretic hormone and angiotensin concentrations, or decrease the approximate-15% falls in estimated plasma volume in the F and NF trials. Nor did fluid ingestion significantly affect the approximate 1.71.h-1 sweat rates, the rises in rectal temperature (from 36.6 degrees to 38.3 degrees C) or the ratings of perceived exertion in the two trials.(ABSTRACT TRUNCATED AT 250 WORDS)