The results of previous epidemiologic research on the possible association between maternal smoking during pregnancy and risk of oral clefts in offspring have been inconsistent. This may be due in part to methodological limitations, including imprecise measurement of tobacco use, failure to consider etiologic heterogeneity among types of oral clefts, and confounding. This analysis, based on a large case-control study, further evaluated the effect of first trimester maternal smoking on oral facial cleft risk by examining the dose-response relationship according to specific cleft type and according to whether or not additional malformations were present. A number of factors, including dietary and supplemental folate intake and family history of clefts, were evaluated as potential confounders and effect modifiers. Data on 3,774 mothers interviewed between 1976 and 1992 by the Slone Epidemiology Unit Birth Defects Study were used. Study subjects were actively ascertained from sites in areas around Boston, Massachusetts and Philadelphia, Pennsylvania; the state of Iowa; and southeastern Ontario, Canada. Cases were infants with isolated defects--cleft lip alone (n = 334), cleft lip and palate (n = 494), or cleft palate alone (n = 244)--and infants with clefts plus (+) additional malformations: cleft lip+ (n = 58), cleft lip and palate+ (n = 140), or cleft palate+ (n = 209). Controls were infants with defects other than clefts, excluding defects possibly associated with maternal cigarette use. There were no associations with maternal smoking for any oral cleft group, except for a positive dose response among infants with cleft lip and palate+ (for light smokers, odds ratio (OR) = 1.09 (95% confidence interval (CI): 0.6, 1.9); for moderate smokers, OR = 1.84 (95% CI: 1.2, 2.9); and for heavy smokers, OR = 1.85 (95% CI: 1.0, 3.5), relative to nonsmokers). This finding may be related to the additional malformations rather than to the cleft itself.