The concept of family history of disease has been used as a surrogate for genetic susceptibility in many epidemiological studies and has also been important as a criterion for selecting individuals for genetic testing. However, little is known about the precise interplay between the true genetic model (genotype-specific penetrances, age of onset distribution), life expectancy, and reproductive patterns in determining the level of family history. In order to address these questions, we performed a simulation study to address these relationships. Factors examined were the age-, sex-, and genotype-specific penetrance of the disease and the distribution of the number of offspring per family. When considering the average number of affected individuals among first-degree relatives of mutation positive probands, penetrance-related factors accounted for 64% of the variance in the average number of affected first-degree relatives, and 58% of the variance in the number of affected first- or second-degree relatives. In general, the average proportion of mutation-positive probands with at least one affected first-degree relative was low, especially for a sex-limited disease, ranging between 20% and 46%, depending on the lifetime penetrance in mutation carriers. Lack of family history among first-degree relatives of mutation positive probands is not necessarily unexpected even for loci conferring relatively high lifetime risk. In selecting probands for genetic testing, we found that under a wide variety of conditions, criteria based on the number of affected among first- and second-degree relatives were superior to those based on first-degree family history alone.