OBJECTIVES: Choline-metabolizing genetic variation may interact with choline intake on fetal programming and pregnancy outcome. This case-control study aims to explore the association of maternal choline consumption and phosphatidylethanolamine N-methyltransferase (PEMT) gene polymorphism rs7946 with incident preterm birth. METHODS: 165 Chinese Han women with preterm delivery and 165 Chinese Han women with term delivery were recruited in Shanghai, China. Dietary choline intake during pregnancy was assessed using a validated food frequency questionnaire. Additionally, DNA samples were genotyped for PEMT rs7946 (G5465A) by a real-time polymerase chain reaction method with plasma homocysteine (Hcy) levels measured by an enzymatic cycling method. Odds ratio (OR) and 95% CI were estimated using unconditional logistic regression with adjustment of the covariates which were selected based on a comparison of social-democratic and clinical characteristics between two groups. RESULTS: There was a significant interaction between maternal choline intake and PEMT rs7946 (P for interaction = 0.04), where the AA genotype carriers who consumed the energy-adjusted choline 255.01 mg/d during pregnancy. Additionally, the highest plasma Hcy was found in the cases with AA genotype and choline consumption