Background: Codeine is the standard antitussive treatment to which novel agents are compared. Little is known about the objective effect of any treatments on cough in chronic obstructive pulmonary disease (COPD). Objective: To quantify the effect of codeine on objective cough frequency (quantified as time spent coughing: cough seconds, cs/h), citric acid cough threshold, and subjective measures in a double-blind, placebo-controlled crossover study in COPD. Methods: We studied 21 patients with physician-diagnosed, stable disease who complained of cough (76.9% male; mean age, 67.7 years; mean predicted FEV 1, 53.4%; median smoking history, 43.5 pack-years). Each subject performed a cough challenge (single breath, citric acid), 10-hour daytime ambulatory and overnight cough recordings, subjective cough scores, and visual analog scales at baseline and on 2 study days, 1 week apart. Codeine phosphate 60 mg or matched placebo were given, in random order, at the start of each cough recording (0 and 12 hours). Results: Median time spent coughing at baseline was 8.27 cs/h (interquartile range [IQR], 5.94-11.67); after placebo treatment, 7.22 cs/h (IQR 4.42-10.40); and after codeine treatment, 6.41 cs/h (IQR 3.86-9.10). Codeine treatment had a significant effect on time spent coughing compared with baseline (P = .02) but not compared with placebo (P = .52). There were no significant differences in cough challenge thresholds (log concentration of tussive agent causing 2 coughs or log concentration of tussive agent causing 5 coughs) or subjective cough measures for codeine compared with placebo. Conclusion: In this study, codeine was no more effective than placebo in patients with COPD complaining of cough. Clinical implications: Codeine is the antitussive agent to which we compare new treatments; however, in a group of stable patients with COPD, it had no effect on cough frequency over placebo. © 2006 American Academy of Allergy, Asthma and Immunology.