A simple and standardized test has been developed to measure airway responsiveness to cold dry air. This consists of stepwise increases in ventilation of dry subfreezing air at 10, 20, 40 and 60% of predicted indirect maximum breathing capacity (IMBC). For each step, the inhalation time was 3 min. The optimal time between the steps was 5 min. Exposure ceased when either a fall in forced expiratory volume in one second (FEV1) of more than 20% of baseline occurred or when there was no response after breathing cold air at 60% predicted IMBC. Moderate isocapnic hyperventilation with cold air beyond 3 min induced no further bronchoconstriction. Varying the interval (0, 2 and 5 min) between the steps produced no significant differences in test results. Changing the pattern of breathing had no effect on airway responsiveness, provided that the patient maintained a constant minute-ventilation. This implies that it is not necessary to monitor the rate and depth of respiration continuously in order to achieve a given minute-ventilation, making the technique simpler. In addition, a "CO2 requirement graph" has been constructed at different levels of ventilation. This allows the inspired CO2 concentration to be preset, eliminating the need for elaborate equipment and monitoring of end-tidal CO2 to keep the subject isocapnic during hyperventilation.