The association of prenatal growth with metabolic syndrome (MS) components and insulin resistance (IR) in children has not been studied in Chile and most developing countries. Some associations found in developed countries are controversial. A retrospective cohort study was designed linking present information on MS components and IR in children with register-based information on birth weight (BW), birth length (BL) and gestational age (GA). Examinations included anthropometry and blood pressure (BP), as well as self-report of pubertal status. A fasting blood sample was taken to determine lipids, glucose, insulin and homeostasis model assessment (HOMA)-IR was calculated. The study cohort of 2152 children was on average 11.4 ± 1.0 years old. The prevalence of MS, IR and overweight were 7.6%, 24.5% and 34%, respectively. Elevated BP was negatively associated with dichotomized risk categories of the perinatal factors studied (BW, BL and GA). Contingency tables showed that high waist circumference (WC) and elevated BP had a U-shaped association with various categories of BW and BL, respectively. Stepwise linear regressions selected: (a) WC as inversely associated to GA and directly associated to BW, (b) BP as inversely associated to GA and (c) HOMA-IR as inversely associated to BL. Non-optimal prenatal growth seems to predispose to high WC, elevated BP and IR in school-age children, supporting the early life origin of several non-communicable diseases. Those associations were rather weak as estimated by the slopes of the regressions and probably reduced by their U-shaped nature; they would reasonably become stronger with a longer follow-up.