OBJECTIVES: Ultrasound-based screening is widely employed for the detection of congenital malformations in utero including congenital heart disease (CHD), but there is widespread variability in the efficacy of screening programs. We aimed to evaluate current antenatal detection rates of selected congenital heart defects in Victoria. METHODS: Data were collected from the Victorian Perinatal Data Collection Unit and Birth Defects Registry. There were 631 209 births in Victoria (1993-2002), of which 4897 cases had CHD. Cases included live births, stillbirths and termination of pregnancies because of CHD. We reviewed all cases from 1999 to 2002 with atrioventricular septal defect, simple coarctation of the aorta, double-inlet or -outlet ventricle, hypoplastic left heart syndrome, simple transposition of the great arteries (TGA), tetralogy of Fallot and truncus arteriosus. Outcome measures were antenatal diagnosis, pregnancy outcome and associated malformations. RESULTS: The overall birth prevalence of CHD from 1993 to 2002 in Victoria was 7.8/1000. The antenatal detection rate for the seven selected defects from 1999 to 2002 was 52.8%. All but 4.8% of the cases had an ultrasound examination at > 13 weeks' gestation. Antenatal detection was highest for hypoplastic left heart syndrome (84.6%) and lowest for simple TGA (17.0%). CONCLUSIONS: This study shows wide variation in the antenatal detection rate of CHD in Victoria. The low antenatal detection rate of TGA, a defect that should be detected easily, demonstrates suboptimal routine obstetric anomaly scanning.