Ventilation and arterial acid-base status at rest and during steady-state work at P i O 2 100, 145 and 250 mm Hg, were studied in: (1) lowlanders at sea level and after 4, 21 and 45 days sojourn at 3100 m altitude; (2) lowlanders residing for 2–15 years at 3100 m; and (3) native altitude residents. The total ventilatory changes at rest and work in the sojourner between ambient conditions at 250 and 3100 m altitudes were attributed: (a) to (acute) hypoxia alone, particularly during moderate to severe exercise; and (b) to the secondary effects of altitude sojourn which accounted for most of the final levels of hyperventilation achieved under all conditions of rest and work, and were complete after 4 days sojourn. Because of: (a) the absence of a significant respiratory alkalosis in response to acute hypoxia at rest, and (b) the normal ventilatory acclimatization to altitude in a lowlander subject who was non-responsive to hypercapnic-hypoxic stimuli combinations, it was reasoned that current theories based on a restoration of CSF [H + ] were not sufficient to explain the hyperventilation obtained upon sojourn to 3100 m. In residents of 3100 m: (a) Exercise e was significantly lower, Pa CO 2 higher (PIo2100), and the Δe with removal of hypoxemia less than in the sojourner at 3100 m, but similar to the sojourner's results obtained during acute hypoxic exposure at sea-level; and (b) native residents and resident low-landers were similar in all respects. Sojourner-resident differences in hypoxic exercise hyperpnea at 3100 m were attributed primarily to the acquired hyper-responsiveness in the sojourner and to a lesser extent to some degree of subnormal ventilatory chemosensitivity in the resident.