ABSTRACT Objective: To compare the incidence of small for gestational age infants among late preterm and term newborns, using the Fenton and Intergrowth-21st curves. Methods: Observational and retrospective study with newborns in a level II maternity. The study was approved by the Institution’s Ethics Committee. Live births from July 2007 to February 2009 with a gestational age from 34 to 41 weeks and seven days were included. Neonates with incomplete data were excluded. Appropriate weight for gestational age was assessed by the Fenton and Intergrowth-21st intrauterine growth curves, considering birth weight <10th percentile as small for gestational age. The degree of agreement between the two curves was assessed by the Kappa coefficient. Numerical variables were compared using the Student t-test or the Mann-Whitney. Categorical variables were compared using the chi-square test. Statistical analyzes were performed using SPSS17® software, considering significant, p<0.05. Results: We included 2849 newborns with a birthweight of 3210±483 g, gestational age of 38.8±1.4 weeks; 51.1% male. The incidence of small for gestational age in the full sample was 13.0 vs. 8.7% (p<0.001, Kappa=0.667) by the Fenton and Intergrowth-21st curves, respectively. Among late preterm, the incidence of small neonates was 11.3 vs. 10.9% (p<0.001; Kappa=0.793) and among full-term infants it was 13.1% vs. 8.5% (p<0.001; Kappa=0.656), respectively for the Fenton and Intergrowth-21st curves. Conclusions: The incidence of small for gestational age newborns was significantly higher using the Fenton curve, with greater agreement between the Fenton and Intergrowth-21st curves among late preterm, compared to full term neonates.