In 14 patients with an oversensitive bronchial system (non-smokers), we investigated the question as to whether quiet breathing in a nitrogen dioxide or sulphur dioxide atmosphere would modify the sensitivity of the airways to sulphur dioxide. On three consecutive days, over a period of 30 minutes at rest, the patients breathed either filtered air or an atmosphere containing 0.25 ppm nitrogen dioxide, or 0.5 ppm sulphur dioxide. There then followed isocapnic hyperventilation of 0.75 ppm sulphur dioxide in increasing ventilation steps of 3 minutes duration each. On three other experimental days, in 7 patients, a stepwise hyperventilation of filtered air was carried out. Quiet breathing of nitrogen dioxide or sulphur dioxide resulted in no obstruction of the airways. The ventilation required to achieve a doubling of the specific airway resistance (PV100SRaw) during hyperventilation of sulphur dioxide was, on average, 46.5, 37.7 and 45.4 l/min after inhalation of filtered air, nitrogen dioxide and sulphur dioxide, respectively. Following nitrogen dioxide, PV100SRaw was significantly smaller (p less than 0.01 than following filtered air or sulphur dioxide. During hyperventilation of filtered air, the average PV100SRaw was 58.2, 51.8 and 55.7 l/min, respectively. We conclude that in non-smokers with an hypersensitive bronchial system, the inhalation of nitrogen dioxide can bring about an increase in the obstructive reaction to sulphur dioxide, without itself leading to an obstruction of the airways; in contrast, sulphur dioxide does not modify the degree of sensitivity.