Objective: To assess the extent of heterogeneity of the genetic risk of age- related macular disease (AMD) among families. Design: Case- controlled population- based familial aggregation study. Participants: Participants comprised 190 first- degree relatives of 65 case probands and 347 relatives of 100 control probands. All probands had been identified from the baseline phase of the Rotterdam Study in The Netherlands. Methods: A family score was computed for each family based on the presence and type of macular disease, the expected risk of disease, and the number, extent of kinship, and age of all family members. Main Outcome Measures: Presence and stage of AMD as diagnosed on fundus photographs, family score, and logistic regression coefficient. Results: The family score of case families showed a peak of approximately 0 with a skewed tail (14% of families) of higher than expected risks of disease toward a maximum of 2.9. The family score of control families centered on 0, apart from 1 outlier. The risk of AMD increased significantly with higher family scores (β =1.34; P< 0.001). Conclusions: The heterogeneity of genetic risk among AMD families is considerable, and the proportion of high- risk families is relatively small. The family score method is relevant for genetic counseling as well as for implementation in studies of genetic dissection of AMD. Objective: To assess the extent of heterogeneity of the genetic risk of age-related macular disease (AMD) among families. Design: Case-controlled population-based familial aggregation study. Participants: Participants comprised 190 first-degree relatives of 65 case probands and 347 relatives of 100 control probands. All probands had been identified from the baseline phase of the Rotterdam Study in The Netherlands. Methods: A family score was computed for each family based on the presence and type of macular disease, the expected risk of disease, and the number, extent of kinship, and age of all family members. Main Outcome Measures: Presence and stage of AMD as diagnosed on fundus photographs, family score, and logistic regression coefficient. Results: The family score of case families showed a peak of approximately 0 with a skewed tail (14% of families) of higher than expected risks of disease toward a maximum of 2.9. The family score of control families centered on 0, apart from 1 outlier. The risk of AMD increased significantly with higher family scores (β = 1.34; P <0.001). Conclusions: The heterogeneity of genetic risk among AMD families is considerable, and the proportion of high- risk families is relatively small. The family score method is relevant for genetic counseling as well as for in research in studies of genetic dissection of AMD.