Previous population-based studies have identified subject characteristics that, when combined, can account for approximately 20% of the observed interindividual variation in baseline SCE rates. In the present investigation, a classic twin study design was used to address the issue of the relevance of genetic factors to baseline SCE rates and to identify other demographic, hematologic, and exposure variables predictive of SCE rate. Questionnaire data and peripheral blood samples from 136 monozygotic and 88 dizygotic twins (age range: 25-81 years) were obtained. Among the large number of variables examined, univariate analyses (including ANOVA tests for the categorical variables and Pearson-product moment correlations for the quantitative variables) revealed smoking status, coffee drinking status, sex, white blood cell count, and absolute numbers of lymphocytes and neutrophils to have significant effects on SCE rates. A stepwise multiple regression analysis showed that together, smoking and coffee drinking status entered at the first step accounted for 21% of the observed variance in SCE, with a further 6% being contributed by the demographic and hematologic variables added in subsequent steps. Finally, the twin analyses showed that after adjustment of the data set for smoking and other significant predictors, genetic factors accounted for approximately 30% of the variation in SCE rates. Thus these data support the hypothesis of a significant genetic influence on baseline SCE.