The purpose of this investigation was to determine if the hyperventilatory response to fatiguing isometric exercise at sea level could predict resting ventilation and acute mountain sickness (AMS) at 4300 m altitude. Exercise consisted of four successive endurance handgrips held to complete fatigue at 40% of maximum isometric handgrip strength (MHS). There was no relationship between the magnitude or pattern of exercise-induced hyperventilation at sea level and the severity of AMS later at altitude. Sea level hyperventilatory response was not predictive of resting ventilation at altitude. Altitude exposure progressively increased both the incidence and magnitude of the hyperventilatory response to exercise and prolonged it for 60-90 s into the recovery period, providing support for the "central command" theory of ventilatory control during isometric exercise. MHS was significantly increased at altitude--by 11% on day 2 and 16% on day 6. Endurance times to fatigue were reduced, but not always significantly so. A follow-up study involving more practice at sea level demonstrated MHS to be significantly increased throughout an entire 18-d stay at 4300 m and for 3, but not 5, days after descent. Significant changes in endurance could not be demonstrated. Neither AMS nor changes in body weight or circulating norepinephrine levels can account for the temporal pattern of increased grip strength, but the respiratory alkalosis occurring at altitude appears to be a likely mechanism.