The biphasic nature of the ventilatory response to sustained hypoxia (< 1 h) is thought to occur as a result of combined effects of stimulation and depression with hypoxia. If the depressant effect of hypoxia on ventilation occurs later than the stimulatory effect, the magnitude of the ventilatory response to progressive hypoxia may be different according to the time required for a given fall in PaO2 during the test. In this study, we measured, in 10 healthy adult volunteers, ventilatory responses to isocapnic progressive hypoxia (HVR) using three different protocols (4 min, 6 min, and 15 min for SaO2 to decrease from the baseline to 80%) with or without pretreatment using theophylline (300 mg twice a day), an adenosine receptor antagonist, in a 2-d, single-blind crossover design. The slope values of HVR with the placebo were 0.31 +/- 0.04 (SE) (L/min/% fall of SaO2) in the 4-min protocol, 0.26 +/- 0.04 in the 6-min protocol, and 0.18 +/- 0.04 in the 15-min protocol. The HVR for the 15-min protocol was significantly lower than that for the 4-min or 6-min protocol. The HVR with theophylline was 0.46 +/- 0.08 for the 4-min protocol, 0.54 +/- 0.09 for the 6-min protocol, and 0.66 +/- 0.11 for the 15-min protocol, among which there were no significant differences. The dependency of HVR on the rate of change in PaO2, that was seen with a placebo run, disappeared after pretreatment with theophylline.(ABSTRACT TRUNCATED AT 250 WORDS)