The role of environmental stimuli in the control of cigarette smoking has been the object of theoretical speculation, but there have been few experimental demonstrations of the control exerted over smoking by environmental events. Self-report data on smoking in the natural environment were compared to observations made of smoking among subjects residing in a laboratory. Participants in the residential studies were exposed to one of two activity schedules in which they were either allowed to determine the duration of each activity (N=11 subjects) or were required to change activities every hour on the hour (N=8 subjects). The smoking patterns of the subjects on the temporally unstructured activity schedule in the residential laboratory were similar to the smoking patterns of subjects in the natural environment. Subjects on the activity schedule showed both changes in the shape of the intercigarette interval distribution and a decrease in the number of cigarettes smoked per day. The findings suggest that smoking patterns may be dependent on temporal characteristics of daily activities and that experimental analyses of smoking patterns may reveal relationships not observed in studies using per-unit time measures of smoking. (Author/JAC)