Maternal smoking during pregnancy is one proposed risk factor for gastroschisis, but reported associations have been modest, suggesting that differences in genetic susceptibility might play a role. We included 108 non‐Hispanic white and 62 Hispanic families who had infants with gastroschisis, and 1,147 non‐Hispanic white and 337 Hispanic families who had liveborn infants with no major structural birth defects (controls) in these analyses. DNA was extracted from buccal cells collected from infants and mothers, and information on periconceptional smoking history was obtained from maternal interviews, as part of the National Birth Defects Prevention Study. We analyzed five polymorphisms in three genes that code for enzymes involved in metabolism of some cigarette smoke constituents (CYP1A1, CYP1A2, and NAT2). Logistic regression models were used to estimate odds ratios (ORs) and 95% confidence intervals (CIs) independently for maternal smoking and maternal and infant gene variants, and to assess joint associations of maternal smoking and maternal or infant gene variants with gastroschisis. In analyses adjusted for maternal age at delivery and stratified by maternal race‐ethnicity, we identified three suggestive associations among 30 potential associations with sufficient numbers to calculate ORs: CYP1A1*2A for non‐Hispanic white mothers who smoked periconceptionally (aOR = 0.38, 95% CI 0.15–0.98), and NAT2*6 for Hispanic non‐smoking mothers (aOR = 2.17, 95% CI 1.12–4.19) and their infants (aOR = 2.11, 95% CI 1.00–4.48). This analysis does not support the occurrence of effect modification between periconceptional maternal smoking and most of the xenobiotic metabolizing enzyme gene variants assessed. © 2014 Wiley Periodicals, Inc.