Identification of the genetic basis for congenital heart defects has been a topic of inquiry for pediatric cardiologists for many years. Any genetic model proposed must account for three often puzzling features of congenital heart diseases, namely the high population incidence, the less than Mendelian recurrence risk, and low concordance rates of heart lesions within individual families. A multifactorial or polygenic inheritance model has been used to explain these observations. Although this model has been serviceable, it does not adequately explain the phenotypic variability within families. Further, the emphasis on teratogenic influences has not led to the elucidation of pathogenetic mechanisms. The multifactorial model should be modified to include the role of chance as originally proposed by Kurnit. Once the role of chance is acknowledged, the phenotypic variability of congenital heart diseases becomes less problematic, and the multifactorial model gives way to a model in which single genes are capable of producing congenital heart disease. Through recent work in animal models and humans, it is now clear that single congenital heart disease genes can be identified that produce phenotypic variability in families that is similar to that seen in the population as a whole. Examples demonstrating recent progress in the search for major congenital heart disease genes are discussed.