The data came from a large, longitudinal cigarette smoking prevention study. Invalid reports occurred when an adolescent reported cigarette smoking at one time point and denied ever trying cigarettes at a subsequent time point. Results revealed that 50 percent of the original sample of 2,597 students reported never having smoked in both seventh and eighth grade, 33 percent reported smoking, 11 percent reported having initiated cigarette smoking in the year between seventh and eighth grade, 4 percent provided misreports of initiation, and 2 percent provided inconsistent responses across the survey. Those with inconsistent responses and those initiating smoking between the two grades were not included in subsequent analyses. Further analysis revealed that misreports did not vary as a function of the complexity of the questionnaire item and did not reflect careless responding. The scores of seventh graders who misreported were higher on the problem-prone variables than were those of nonsmoker and lower than those of smokers. In contrast, when measured in the eighth grade, misreporters' scores did not differ from those of nonsmokers. Misreporters did display a greater decrease in intentions to smoke, alcohol use, and number of friends who smoked compared to nonsmokers and consistent smokers. Tables, note, and 14 references (Author abstract modified)