For simple Mendelian traits the relative risks In a certain type of relative can be determined when the phenotype (or genotype) of the proband is known. On the other hand, for a quantitative trait the relative contribution of gene and environment is determined by assembling data on correlation between relatives. When the trait is purely qualitative but the age of onset varies, the concept of correlation between relatives can be extended to the joi nt distribution of the age of onset in probands and their relatives. This approach serves two purposes: first, through this we can predict the age-specific risk rates for relatives of a proband whose age of onset for the trait is known; and second, if the parametric form of the hazard rate in the general population is known, the resolution of genetic versus nongenetic causes of the trait can be attempted from these studies. Here we outline the principles of the joint distribution of age of onset among relatives in general terms, and illustrate the statistical power of the method to detect certain changes in the parameters of the hazard function which indirectly indicate the relative influences of genetic versus nongenetic factors to the trait. The merits of this approach over the other forms of pedigree analysis are also discussed.