We tested the hypothesis that growth retardation in symmetric small-for-gestational-age (SGA) infants may start in the first trimester of pregnancy, whereas in asymmetric SGA infants, it may start in the third trimester. We also examined if there may be a brain sparing effect in asymmetric SGA infants. Infants were classified as symmetric or asymmetric SGA infants, and 474 non-SGA infants. Intrauterine growth was assessed by prenatal ultrasonic measurements of fetal abdominal diameter, femur length and biparietal diameter at week 17, 25, 33 and 37 of gestation. Cross-sectional data as well as longitudinal growth curves suggested that growth retardation in both SGA groups started in the second trimester, and followed similar patterns until birth. Thus, our results did not support the hypothesis that symmetric and asymmetric growth retardation is associated with clear temporal differences in growth, and we found no sign of brain sparing in asymmetric compared to symmetric SGA infants.