Background: The extent of exposure to tobacco toxicants in smokers who have reduced their cigarette intake compared with smokers who are light smokers is relatively unknown. The goal of this study is to investigate the occurrence of compensatory smoking in reducers compared with light smokers by measuring toxicant exposure. Methods: Participants in two smoking reduction intervention studies (N = 64) were selected for comparison with a group of light smokers (N = 62) who smoked the same number of cigarettes as the reducers. A compensatory smoking score was defined (biomarker level for reducer/biomarker level for light smoker) and calculated for urinary 4-(methylnitrosamino)-1-(3-pyridyl)-1- butanol (NNAL) and its glucuronides (total NNAL), metabolites of the tobacco-specific lung carcinogen 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone, to measure the degree of smoking compensation in reducers when compared with the light smokers. Results: The mean level of creatinine-adjusted total NNAL for reducers was over twice that of light smokers even when they smoked about the same number of cigarettes per day. The difference of the mean total NNAL concentrations between light smokers and reducers was highly significant (P < 0.0001). Wide variability in total NNAL concentrations was also observed in reducers, with the extent of this variability between light smokers and reducers being significantly different (P = 0.0005). The level of individual reduction was shown to be a consistent predictor of compensatory smoking (r = 0.50; adjusted Ps = 0.002), with greater cigarette reduction associated with more compensation. Conclusions: Compensatory smoking limits the harm reduction value of decreased smoking of cigarettes.