A three-study series was conducted to examine smoking topography across clinical, laboratory, and naturalistic settings for 24 smokers who were patients in an alcohol treatment program. Experiments 1 and 2 included surreptitious observation of smoking in group therapy sessions (naturalistic), and obtrusive observation of smoking in a smoking clinic setting. The third study added a laboratory smoking condition to the clinic and group conditions. Puffing patterns in clinical and laboratory settings tended to be similar, and both differed from smoking in the naturalistic setting. Subjects generally took more puffs, longer puffs, and had shorter cigarette durations in clinical and laboratory settings than in the naturalistic setting. The implications of these findings for assessment of intake and health risk are discussed.