Gestational diabetes mellitus (GDM) is associated with adverse perinatal outcomes because of suboptimal glucose management and glucose control and excessive weight gain. Metformin can offset these factors but is associated with small for gestational age (SGA) infants. We sought to identify risk factors for SGA infants, including the effect of metformin exposure on SGA status. In this prespecified secondary analysis of the EMERGE trial, which evaluated the effectiveness of metformin vs placebo in treating GDM and found reduced gestational weight gain and longer time to insulin initiation with metformin use, we included women with a live-born infant and known infant birthweight and gestational age at delivery. We compared the numbers of SGA infants in both groups and explored baseline predictive factors to help identify those at highest risk of delivering an SGA infant. Baseline maternal characteristics were similar between SGA and non-SGA pregnancies. On multivariable-adjusted regression, no baseline maternal variables were associated with SGA status. Mothers of SGA infants were more likely to develop pre-eclampsia or gestational hypertension (18.2% vs 2.0%, p=0.001; 22.7% vs 5.4%, p=0.005, respectively); after multivariable adjustment, pre-eclampsia was positively associated with SGA status). Among SGA pregnancies, important perinatal outcomes including preterm birth, Caesarean delivery and neonatal care unit admission did not differ between the metformin and placebo groups (20.0% vs 14.3%, p=1.00; 50.0% vs 28.6%, p=0.25; 13.3% vs 42.9%, p=0.27, respectively). Pre-eclampsia was strongly associated with SGA infants. Metformin-exposed SGA infants did not display a more severe SGA phenotype than infants treated with placebo. Clinical Trials.gov NCT02980276; EudraCT number: 2016-001644-19.