We investigated the effects of nightly intermittent exposure to hypoxia and of training during intermittent hypoxia on both erythropoiesis and running economy (RE), which is indicated by the oxygen cost during running at submaximal speeds. Twenty-five college long- and middle- distance runners [maximal oxygen uptake (Vo(2max)) 60.3 +/- 4.7 ml x kg(-1) x min(-1)] were randomly assigned to one of three groups: hypoxic residential group (HypR, 11 h/night at 3,000 m simulated altitude), hypoxic training group (HypT), or control group (Con), for an intervention of 29 nights. All subjects trained in Tokyo (altitude of 60 m) but HypT had additional high-intensity treadmill running for 30 min at 3,000 m simulated altitude on 12 days during the night intervention. Vo(2) was measured at standing rest during four submaximal speeds (12, 14, 16, and 18 km/h) and during a maximal stage to volitional exhaustion on a treadmill. Total hemoglobin mass (THb) was measured by carbon monoxide rebreathing. There were no significant changes in Vo(2max), THb, and the time to exhaustion in all three groups after the intervention. Nevertheless, HypR showed approximately 5% improvement of RE in normoxia (P < 0.01) after the intervention, reflected by reduced Vo(2) at 18 km/h and the decreased regression slope fitted to Vo(2) measured during rest position and the four submaximal speeds (P < 0.05), whereas no significant corresponding changes were found in HypT and Con. We concluded that our dose of intermittent hypoxia (3,000 m for approximately 11 h/night for 29 nights) was insufficient to enhance erythropoiesis or Vo(2max), but improved the RE at race speed of college runners.