To investigate the effect of altitude exposure on running economy (RE), 22 elite distance runners [maximal O(2) consumption (Vo(2)) 72.8 +/- 4.4 ml x kg(-1) x min(-1); training volume 128 +/- 27 km/wk], who were homogenous for maximal Vo(2) and training, were assigned to one of three groups: live high (simulated altitude of 2,000-3,100 m)-train low (LHTL; natural altitude of 600 m), live moderate-train moderate (LMTM; natural altitude of 1,500-2,000 m), or live low-train low (LLTL; natural altitude of 600 m) for a period of 20 days. RE was assessed during three submaximal treadmill runs at 14, 16, and 18 km/h before and at the completion of each intervention. Vo(2), minute ventilation (Ve), respiratory exchange ratio, heart rate, and blood lactate concentration were determined during the final 60 s of each run, whereas hemoglobin mass (Hb(mass)) was measured on a separate occasion. All testing was performed under normoxic conditions at approximately 600 m. Vo(2) (l/min) averaged across the three submaximal running speeds was 3.3% lower (P = 0.005) after LHTL compared with either LMTM or LLTL. Ve, respiratory exchange ratio, heart rate, and Hb(mass) were not significantly different after the three interventions. There was no evidence of an increase in lactate concentration after the LHTL intervention, suggesting that the lower aerobic cost of running was not attributable to an increased anaerobic energy contribution. Furthermore, the improved RE could not be explained by a decrease in Ve or by preferential use of carbohydrate as a metabolic substrate, nor was it related to any change in Hb(mass). We conclude that 20 days of LHTL at simulated altitude improved the RE of elite distance runners.