To examine the changes in birth cohort prevalence rates and severity of congenital heart disease, we studied children with congenital heart disease born to blacks, whites, and Mexican-Americans in Dallas County from 1971 through 1984. Diagnoses were made by pediatric cardiologists' clinical evaluations, echocardiography, catheterization, surgery, or autopsy. During this study period, 2,509 of 379,561 liveborn infants were diagnosed, a prevalence rate of 6.6/1000. The rates for whites was significantly higher than for blacks or Mexican-Americans -7.2/1,000, 5.6/1,000, and 5.911,000, respectively. The rate for severe cases requiring cardiac catheterization or surgery or undergoing autopsy was 3.111,000 and did not differ among the three groups. The time trend for rates of congenital heart disease suggested an apparent increase in prevalence rate during the 1970s; however, the prevalence rate of severe forms remained relatively stable. This indicates that the apparent rise in prevalence could be accounted for by an increase in detection of mild cases. These findings were interpreted as reflecting a greater tendency for pediatricians to refer asymptomatic children with significant heart murmurs to a pediatric cardiologist. (Circulation 1990;81:137-142) C ongenital heart disease (CHD) is a leading cause of death during the first year of life. Malformations of the cardiovascular system are also associated with significant medical morbidity, which requires use of costly medical facilities. Children with CHD use 25-30% of the beds in most large pediatric intensive care units and, therefore, consume a large fraction of pediatric health care resources. To plan future health care needs, epidemiologic data are needed regarding changes in prevalence and severity of CHD. The surveillance systems for birth defects conducted by the Centers for Disease Control have indicated a more than twofold increase in the most common form of CHD, ventricular septal defect, between-1970 and 1977.1 To examine the changes in prevalence rates and severity of CHD, we gathered epidemiologic data on children with CHD who were born in a well-defined geographic area during a specific time period. The