A diuretic drug (40 mg of furosemide) was utilized to study the effects of dehydration (D) on competitive running performance, without prior thermal or exercise stress. Eight men competed in randomized races of 1,500, 5,000, and 10,000 m, while normally hydrated (H) and with mean plasma volume reductions of 9.9, 12.3, and 9.9%, respectively. As a result of the reduced body water (change in body weight = -1.9, -1.6, and -2.1%), mean outdoor performance times on a running track increased 0.16 min, 1.31 min (P less than 0.05), and 2.62 min (P less than 0.05) in the 1,500-m, 5,000-m, and 10,000-m trials. Running performance decrements due to dehydration were more strongly correlated with changes in body weight (r = -0.79, -0.65, and -0.40) than with urine volume or plasma volume differences. In addition, subjects were studied during submaximal and maximal treadmill exercise while H and D (mean change in plasma volume = -7.1%). Neither submaximal nor maximal oxygen uptake was significantly altered (P greater than 0.05) as a consequence of D. Mean treadmill run time to volitional exhaustion was reduced by 41.4 s (P less than 0.05) during the D treadmill trial. Therefore, it appears that competitive performance in trials of long duration (5,000 and 10,000 m) was affected to a greater extent by D than the shorter 1,500-m event, even though submaximal and maximal oxygen uptake was not altered.