To control for likely overreporting of abstinence in clinical trials of smoking cessation aids, field convention demands the corroboration of subjects’ self-reports by a biochemical/pharmacological marker. It is, however, currently debated if urinary cotinine (UC), a metabolite of nicotine, should be preferred because of its higher sensitivity, although sample collection for and analysis of cotinine are much more expensive and work intensive than carbon monoxide (CO) measurements in exhaled air. In the present study, it turned out that UC was of only moderately higher sensitivity than CO (99.4% vs. 96.3%; p = 0.02), the difference being significant only at group sizes of >164. UC identified participants as smokers who escaped CO detection in 4.9% of the cases, whereas CO identified smokers who escaped UC detection in 2.7% of the cases (p = 0.014). Our findings suggest that the costs/disadvantages of using UC instead of CO may outweigh its benefit as a pharmacological marker of (non)smoking status.