Objective To measure the association of fast food density (FFD) and supermarket density (SD) with adverse neonatal outcomes in pregnancies with gestational diabetes mellitus (GDM). Study Design This was a retrospective cohort study of women with GDM who delivered at a tertiary care center in a large metropolitan area (1/2010-2/2016). ZIP codes were used to link women with surrounding food environment. FFD and SD were calculated as the number of establishments per 100,000 residents for each ZIP code and classified into quartiles. Quartile 1 represented lowest FFD or SD. Four neonatal outcomes were assessed: large for gestational age (LGA) neonate, shoulder dystocia, neonatal hypoglycemia, and neonatal intensive care unit admission > 3 days. Bivariable and multivariable analyses estimated associations of FFD and SD with outcomes. Results A total of 2,373 women met eligibility criteria. Insurance, race/ethnicity, language, and body mass index differed between quartiles, although GDM type did not. High FFD was associated with lower odds of LGA but not with other outcomes. Low SD and FFD:SD ratios were not associated with any outcomes. Conclusion Among women with GDM, high FFD was associated with some neonatal outcomes, but low SD and FFD:SD ratio was not. Food environment may not be the only social determinant of neonatal outcomes in pregnancies with GDM.