Ten Huntington disease (HD) families are analyzed using maximum likelihood methods to study age-of-onset (AO) heterogeneity. Both age of onset and age at examination are used in calculating an individual's likelihood of being affected; familial correlations and family structure are not included in the model. The model allows for differences in current age distributions by considering the distribution of age of onset conditional on age at examination. Families are grouped according to family type (ie, juvenile-onset families and adult-onset families) and then analyzed within each family type as well as between the two types, using both a chi-square test for heterogeneity and F-ratios. The results of the chi-square analyses indicate heterogeneity in AO among the individual families as well as between the two family types. However, the results of the F-ratios show no significant difference between the two family types, indicating that the difference in the chi-square analysis between the two family types is mostly owing to variation among the families. Thus, the results demonstrate that the existing variation in age of onset between these two family types is dwarfed by the magnitude of the variation among the families within each type. This implies that families of sufficient size should be evaluated individually for age-of-onset data for the purpose of either genetic counseling or linkage studies.