Data from a case-control study of childhood acute myeloid leukemia (AML) including 187 matched case-control pairs were examined for evidence of associations between parental cigarette smoking and alcohol consumption and the subsequent development of childhood AML. The cases were stratified by French-American-British morphology in order to evaluate potential differences in risk based on this classification system. There was little evidence of any association between cigarette smoking by parents during the index pregnancy and childhood AML. There was some evidence of an increased risk of AML among children who were diagnosed at or before 2 years of age and whose mothers reported consuming alcohol during their pregnancies (odds ratio, 3.00; 95% confidence interval, 1.23 to 8.35). This finding appeared to be especially pronounced for AML with a monocytic component (M4/M5) (odds ratio, 9.00; 95% confidence interval, 1.25 to 394.5), but a cautious interpretation of these data are advised because of the small number of subjects included in this subgroup analysis. Since this is one of the first case-control studies to evaluate specific French-American-British subtypes of AML, these results may be viewed as generating hypotheses. It is suggested that future studies of childhood AML include: (a) a sufficient number of cases to permit an evaluation of the morphological classification of the cases; and (b) additional questions on the alcohol consumption of the mother during the time of her pregnancy with the index child.