Multivariate survival models are shown to be appropriate for the analysis of the genetic and the environmental nature of a human life-span. Models which involve continuously distributed individual frailty, play an important role in the genetic analysis of an individual's susceptibility to disease and death. These models, however, are not appropriate for the detection of the effects of separate genes on survival. For this purpose we developed a 'major gene' frailty model of multivariate survival and applied it to simulated and real pedigree data. The analysis shows that this model can be used for the detection of the presence of major genes in the population and for the evaluation of the effects of such genes on survival.