OBJECTIVE To compare antenatal diagnoses with autopsy findings in pregnancies terminated after ultrasound detection of fetal anomalies. A second aim was to study the quality of antenatal fetal diagnosis over time. METHODS Retrospective, multicenter study over two consecutive six-year periods in Uppsala and Stockholm. METHODS Cases were identified through fetal autopsy reports. METHODS Three hundred and twenty-eight fetuses from pregnancies terminated between 1992 and 2003 because of ultrasonographically diagnosed anomalies. METHODS The findings at the last ultrasound examination were compared with the autopsy reports. RESULTS In 299 cases (91.2%) ultrasound findings either exactly matched or were essentially similar to the autopsy findings. In 23 cases (7%) ultrasound findings were not confirmed at autopsy, but the postnatal findings were at least as severe as the antenatal ones. In six cases (1.8%) termination was performed for an anomaly which proved to be less severe than was predicted by ultrasound. The number of such cases was the same in both six-year periods, while the total number of cases increased from 113 in the first to 215 in the second period. Fetal examination provided further diagnostic information in 47% of the cases. In 10% a syndrome was disclosed. CONCLUSIONS Termination of pregnancy was not always based on a correct antenatal diagnosis. All fetuses but one from terminated pregnancies had evident anomalies. In six cases (1.8%) the decision to terminate was based on suboptimal prognostic and diagnostic information. Fetal autopsy by an experienced perinatal pathologist is essential to provide a definitive diagnosis.