OBJECTIVE To determine in low birth weight infants the relations of being small for gestational age at birth, microcephalic at birth, low weight for age at 2 years, and microcephalic at 2 years to full scale intelligence quotient (FSIQ) at the age of 16 years. METHODS A prospective observational study of 422 of 837 eligible nondisabled low birth weight (<2000 g) adolescents from the newborn brain hemorrhage cohort with weight and head circumference at birth and at the age of 2 years in whom FSIQ was assessed with the Wechsler Abbreviated Scales of Intelligence at the age of 16 years. RESULTS In a multiple regression analysis, being small for gestational age (beta = -0.14, p = .02) and microcephalic at 2 years (beta = -0.18, p < .001), but not low weight for age at 2 years or microcephaly at birth, had significant independent effects on 16-year FSIQ. After adjusting for pre-, peri-, and postnatal risk factors for poor cognitive performance, the independent effects of being small for gestational age (beta = -0.13, p = .004) and microcephalic at 2 years (beta = -0.13, p = .01) persisted. In this analysis, birth social risk had the largest significant independent effect on 16-year FSIQ (beta = -0.28, p < .001). CONCLUSIONS Being small for gestational age at birth, but not low weight for age at 2 years, and microcephaly at 2 years, but not at birth, were independently related to FSIQ of nondisabled low birth weight adolescents, both with and without control for pre-, peri-, and postnatal risk factors.