The phenomenon of senescence requires both evolutionary and proximate explanations. The most widely accepted evolutionary explanation for senescence is that it never gets exposed to natural selection because environmental hazards kill all individuals before the age at which senescence causes decreased fitness. If this explanation is sufficient, wild populations should not demonstrate senescence, and their mortality rates should therefore remain constant during adult life, except when environmental causes of mortality have recently decreased. The alternative explanation for the persistence of the genes that cause senescence is that they have been selected for because they have pleiotropic effects that are beneficial early in life when the force of selection is strongest. Where this is the case, mortality rates should increase with age in wild populations. A method is described for using life table data to calculate an estimate of the intensity of selection acting on senescence in wild populations. This method is applied to a variety of life tables. The results suggest that pleiotropic genes may be important causes of senescence in some populations, but not in others. This has implications for research on the proximate mechanisms of senescence.