The effects of three types of hyperventilation challenge tests (free hypocapnic, ambient air; controlled isocapnic, ambient air; controlled isocapnic, dry air), on FEV1 and on Raw were compared in 12 asymptomatic asthmatics. Controlled isocapnic hyperventilation of dry air as well as of ambient air caused bronchoconstriction lasting for more than 8 min, but the degree of bronchoconstriction was significantly greater with the former. Free hypocapnic hyperventilation of ambient air caused peak bronchoconstriction after 1 min, followed by a steady functional improvement; the values after 1 min were similar to those following isocapnic dry air hyperventilation, and values after 8 and 15 min were similar to those following isocapnic ambient air hyperventilation. The changes in Raw induced by the three types of hyperventilation were not influenced by a preceding full inspiration. In the eight subjects in whom the lung function had returned to within 10% of control after 30 min, identical duplicate hyperventilation challenge tests performed at that time demonstrated a significant, partial refractoriness for all tests. Thus the simple, free hypocapnic, hyperventilation test with ambient air was found to be as reliable as the more sophisticated controlled isocapnic tests.