BACKGROUND: Although both fetal and long-term growth restriction are well documented in fetal alcohol spectrum disorders, effects on pattern of growth trajectory have not been characterized. Furthermore, the degree to which growth trajectories are related to fetal alcohol-related neurocognitive deficits is unknown. METHODS: Ninety-three heavy drinking pregnant women and 64 controls were recruited at initiation of prenatal care in Cape Town, South Africa. Small for gestational age (SGA) was defined as birth weight <10th percentile. Length/height, weight, and head circumference were measured at 6.5 and 12 months and 5, 9, and 13 years. Four growth trajectories were identified: SGA with long-term postnatal growth restriction (length/height-for-age <10th percentile through 13 years); SGA with catch-up growth; no SGA or postnatal growth restriction; and late-onset postnatal stunting. IQ was assessed at 5 and 10 years, and learning, memory, and executive function at 10 years. RESULTS: Children born SGA with postnatal growth restriction were most heavily exposed. Exposure was intermediate for those born SGA with catch-up growth and lowest for those without prenatal or postnatal growth restriction. Effects on neurocognition were strongest in children with both prenatal and long-term growth restriction, more moderate in those with fetal growth restriction and postnatal catch-up, and weakest in those without growth restriction. CONCLUSIONS: These findings validate the use of growth restriction in the diagnosis of fetal alcohol spectrum disorders and identify growth trajectory as a biomarker of which heavily exposed children are at greatest risk for cognitive developmental deficits.