A systematic study was undertaken to estimate the birth prevalence of congenital heart disease because there was a clinical impression that a disproportionate number of cases occurred in Yuma, Arizona. Control data were obtained from Sierra Vista, Arizona, a region with similar demographic characteristics, and from the Baltimore-Washington Infant Study. Patients with chromosomal or syndromal associations were excluded. In the Baltimore-Washington Infant Study only echocardiographic and invasively documented cases were included. From 1983 to 1988 the birth prevalence was significantly higher in Yuma (10.5/1,000) than in Sierra Vista (5.4/1,000, p less than 0.0001). As assessed only by invasive or echocardiographic diagnosis, there was a higher birth prevalence of congenital heart disease in the study population (6.7/1,000) compared with both the Baltimore-Washington Infant Study (3.7/1,000, p = 0.0008) and Sierra Vista (4.6/1,000, p = 0.04). Families were interviewed to exclude cases in which the mother did not spend the month before conception and the first trimester in Yuma or Sierra Vista. The birth prevalence for Yuma (6.0/1,000) remained significantly greater than that for Sierra Vista (3.8/1,000, p = 0.03). The exclusion of cases in which the mother resided elsewhere suggests, but does not prove, that an environmental influence may have played a role in the increased birth prevalence of congenital heart disease in this community.