Objectives To evaluate in a population-based cohort the effect of the introduction of the 20-week ultrasound scan in 2007 on the time of diagnosis, pregnancy outcome and total prevalence and liveborn prevalence of cases with selected congenital heart defects (CHDs) in The Netherlands. Methods We included children and fetuses diagnosed with selected severe CHD, born in the 11-year period from 2001 to 2011. Two groups of CHD were defined: those associated with an abnormal four-chamber view at ultrasound (Group 1), and those associated with a normal four-chamber view at ultrasound (Group 2). The time of diagnosis, pregnancy outcome and total liveborn prevalence were compared for both groups over two 5-year periods, before and after the introduction of the 20-week ultrasound scan. Trends in total and liveborn prevalence were examined over 2001 to 2011. Results Information was collected on 269 children and fetuses. After the introduction of the 20-week ultrasound scan, the prenatal detection rate of CHDs increased in both groups (Group 1, 34.6% in 2001-2005 vs 84.8% in 2007-2011 (P < 0.001); Group 2, 14.3% in 2001-2005 vs 29.6% in 2007-2011 (P = 0.037)). The rate of termination of pregnancy (TOP) increased significantly only for Group 1 (15.4% vs 51.5% (P < 0.001)). The total prevalence of CHD in Group 1 increased over time from 2.9 per 10 000 births in 2001 to 6.4 per 10 000 births in 2011 (P = 0.016). The liveborn prevalence did not show a trend over time. For Group 2, no trends in total or liveborn prevalence could be detected over time. Conclusions Since the implementation of the routine 20-week ultrasound scan in The Netherlands, prenatal detection rate of selected severe CHDs increased significantly. Improved prenatal detection was accompanied by a more than three-fold increase in TOP, although only in those CHDs with an abnormal four-chamber view at prenatal ultrasound.