It has been reported that short-term treatment with relatively high doses of opiates or promethazine causes improvements in dyspnoea and exercise tolerance in patients with chronic airflow obstruction (CAO). This study was designed to determine whether initial benefits were sustained during chronic administration of codeine or promethazine and to compare the two drugs in terms of their efficacy and possible mechanisms of action. Eleven patients with stable CAO were entered into a double-blind, randomized cross-over trial in which codeine (30 mg four times daily) or promethazine (25 mg four times daily) were orally administered for 1-month periods. Treatment effects were assessed by spirometry, arterial blood gases, 12-minute walk distance and subjective dyspnoea ratings. A statistically significant increase from the baseline in mean arterial PCO2 at at 24 hours (P less than 0.01) and at 1 month (P less than 0.05) occurred with codeine administration. There was no significant change from baseline for any other measurement with either drug, and no differences were detected between the two treatment arms. Four of the eleven patients did not complete the study; three of the four experienced worsening of their CAO requiring hospitalization (two while receiving codeine, one while receiving promethazine). We conclude that chronic treatment with either codeine or promethazine provides uncertain benefits to patients with CAO which may not outweigh potential risks.