A population-based case-control study of acute non-lymphocytic leukemia (ANLL) was performed with 80 ANLL cases diagnosed between 1973 and 1979, who were derived from the nationwide register of the Dutch Childhood Leukemia Study Group. Cases were compared to three age- and sex-matched population controls and, in order to control for recall bias, to 517 cases with acute lymphocytic leukemia from the same study base. Information on a large number of exposures to putative risk factors was collected by a self-administered questionnaire mailed to the parents. No significant association of ANLL was observed with smoking habits of the mother during pregnancy, ultrasound examinations, prenatal exposure to x-rays, viral infections, or hydrocarbon exposure. When comparing ANLL cases to population controls, maternal use of alcohol during pregnancy was associated with a more than two-fold increased risk of ANLL (odds ratio = 2.6; 95% confidence interval = 1.4-4.6). A similar increase in risk was found when comparing ANLL cases to acute lymphocytic leukemia cases. There was no significant elevation in risk for ANLL found for parental use of alcohol 1 year before pregnancy. This study suggests that intrauterine exposure to alcohol may increase the risk for childhood ANLL.