Cigarette smokers have been reported to void urine that is more mutagenic, as measured in the Salmonella/microsome assay, than urine voided by nonsmokers. Several previous studies have attempted to correlate indices of tobacco smoke exposure (e.g. nicotine, cotinine, tar intake) with urinary mutagenicity, with conflicting results. These studies generally involved small numbers of smokers and did not carefully control diet, which is known to affect urinary mutagenicity markedly. Our objective was to conduct a controlled study to determine clearly if there were a correlation between urinary nicotine, cotinine, or nicotine + cotinine and urinary mutagenicity and to determine if nicotine or its major metabolite plays a role in the mutagenicity of urine from cigarette smokers. We used a large number of smokers (31), each of whom smoked both a tobacco-burning cigarette and a tobacco-heating cigarette on consecutive weeks, and we prepared and served identical diets to all subjects. Nicotine and cotinine concentrations were determined in small aliquots from urine samples collected over 24 hr, and the remaining urine sample was extracted and concentrated on XAD-2 resin for mutagenicity assays in the Salmonella/microsome test. Nicotine, cotinine, and nicotine + cotinine were statistically correlated with mutagenicity of urine from smokers of the tobacco-burning cigarette, but there was no correlation between nicotine, cotinine, or nicotine + cotinine and mutagenicity of urine from smokers of the tobacco-heating cigarettes. Thus, although urinary nicotine and cotinine concentrations correlate with urinary mutagenicity in smokers of tobacco-burning cigarettes, the present results indicate that nicotine and its metabolite are not responsible for the mutagenicity of smokers' urine.