To identify maternal/pregnancy characteristics, first trimester ultrasound parameters and biochemical indices which are significant independent predictors of small-for-gestational age (SGA) and large-for-gestational age (LGA) neonates. Design. Retrospective cross-sectional study. Setting. Two fetal Medicine Units. Population. 4 702 singleton pregnancies presenting for screening for chromosomal abnormalities by nuchal translucency and maternal serum biochemistry at 11–14 weeks. Methods. Reference ranges for birthweight applied to our population were constructed by the Royston and Wright method. Multiple logistic regression was applied to develop first trimester prediction models for SGA and LGA. Main outcome measures. Birth of SGA or LGA neonate. Results. Maternal height, parity, smoking, assisted conception, delta crown–rump length, delta nuchal translucency, free beta human chorionic gonadotrophin and pregnancy-associated plasma protein-A were significant independent predictors of SGA. Maternal weight and height, smoking, delta crown–rump length and delta nuchal translucency were significant independent predictors of LGA. Models for SGA (AUC=0.7296, CI: 0.69–0.76, p<0.0001) and LGA (AUC=0.6901, CI: 0.65–0.72, p<0.0001) were derived, applicable to routine obstetric population at low risk for these conditions. For 20% screen positive rate the modeling achieves sensitivities of about 55% for SGA and 48% for LGA neonates. Conclusion. Prediction for birthweight deviations is feasible using data available at the routine 11–14 weeks' examination. Delta CRL and delta nuchal translucency were significant independent predictors for both SGA and LGA.