Iron-deficiency anemia impairs exercise capacity, but whether nonanemic iron depletion decreases endurance performance is unclear. In 14 iron-deficient (serum ferritin level, less than 20 micrograms/L [less than 20 ng/L])nonanemic runners, hematologic and treadmill running values were followed up during a competitive season. Following a four-week control period, runners were treated for one month in a double-blind protocol with ferrous sulfate (975 mg/d) or placebo. During treatment, the mean ferritin level rose from 8.7 to 26.6 micrograms/L (8.7 to 26.6 ng/mL) in those patients taking iron and fell from 10.6 to 8.6 micrograms/L (10.7 to 8.6 ng/mL) in the placebo group. Treadmill endurance times improved significantly in the iron-treated runners compared with controls. Endurance time declined in all seven controls (range, 0.07 to 1.30 minutes), while six of seven iron-treated subjects improved their performance (range, 0.03 to 1.92 minutes). No significant differences in maximal or submaximal oxygen consumption, ventilation, or heart rate were observed between the groups except for a 4% increase in maximum oxygen consumption during placebo treatment. These data indicate that nonanemic iron deficiency impairs exercise performance but does not influence gas exchange or cardiac measures.