Objective To assess prenatal heart disease screening program by ultrasound. Methods A total of 11 544 second-trimester screening scans were performed before 24 weeks′ gestation on 11 410 women between February 2004 and May 2007 in Obstetrics and Gynecology Hospital of Fudan University. Fetal heart screening was based on four-chamber and outflow tract views( left ventricular outflow + three vessel view). The sensitivity and specificity of different views were evaluated. Follow-up data of newborns was obstained. Results (1) Among 11 544 cases, 48 cases of congenital heart disease (CHD) were diagnosed in utero. Six cases were false negative, and 2 cases were false positive. The incidence of CHD was 0.47% (54/11 544). (2) Thirty-three CHDs were detected based on the four-chamber view, including 18 ventricular septal defect (9 with conotroncal anomalies),6 anomalous atrioventricle valve, 9 disproportion of left/right ventricle. The sensitivity of the four-chamber view alone was 61.11% (33/54), and the specificity was 99. 98% ( 11 488/11 490). Fifteen CHDs were detected based on the left ventricular outflow and three vessel views, including 1 pulmonary atresia, 3 pulmonary valve stenosis, 2 transposition of the great arteries ( TGA ), 1 pulmonary stanosis with TGA, 6 tetralogy of Fallot, and 2 pulmonary stenosis. The sensitivity of the combination of the four-chamber and outflow tract views was 88. 89% (48/54), and the specificity was 99. 98% (11 488/11 490). (3)Of 48 CHDs,11 cases were eceompanied by other malformations. Eleven cases were performed amniocentesis, amony whom 5 cases were trisemy 21. Conclusion The screening program based on four-chamber and outflow tract views shows good sensitivity and excellent specificity. Our prenatal heart screening program is clinically feasible.