OBJECTIVE To ascertain whether the risk for congenital malformations is increased in pregnancies with deviating fetal growth, i.e. in those in which the estimated date of delivery (EDD) was postponed more than 1 week at the second-trimester ultrasound fetometry scan, or those suspected of intrauterine growth restriction at routine ultrasound fetometry in the third trimester. METHODS The study period was 1994-2003. We used the Swedish Medical Birth Register to identify pregnancies with appropriate data (n = 605 845). A regional ultrasound database consisting of 73 092 pregnancies was used for more detailed data and analysis of the third-trimester fetal growth. The number of congenital malformations was ascertained from three national health registers. RESULTS We found a moderately increased risk for any malformation in the group of fetuses in which the EDD was postponed > 1 week. The strongest effect was seen for chromosome anomalies and central nervous system malformations, including neural tube defects. In the third trimester, an increased risk for fetal malformations was found in asymmetrically vs. symmetrically growth-restricted fetuses, perhaps more strongly when adjustment of the EDD had been done at the dating scan. CONCLUSIONS Fetuses in which the EDD differs between that calculated by the last menstrual period and that calculated by second-trimester ultrasound measurement seem to have an increased risk for congenital malformations, including chromosomal anomalies. A targeted ultrasound examination for malformation screening might be recommended for this group. A similar policy might be recommended when intrauterine growth restriction, especially of the asymmetrical type, is suspected later in pregnancy.