In an attempt to estimate the onset of aging-associated mortality in humans, Swiss national survival and mortality data from 1978 to 1983 were analyzed. The nonparametric kernel method served to estimate gender-specific survival and hazard functions related to five major as well as all causes of death. On the basis of graphical models, it was hypothesized that the onset of aging conceivably was associated with prominent acceleration in mortality rates. The earliest maximal accelerations in hazard functions for most causes of death occurred during the second and again during the fourth age decade. Overall mortality rates of males and females exhibited prominent humps between the two periods of acceleration. These humps were accounted for largely by a high incidence of deaths from violence (accidents, suicide, and crime), which have to be attributed to environmental factors rather than to senescence. On the other hand, no plausible argument could be found against the assumption that maximal acceleration in death rates from ischemic heart and other circulatory diseases around 20 years of age was related to aging. Therefore, these data were interpreted to indicate that in the population examined, senescent mortality sets in around 20 years of age, about 5 years earlier in males than in females. However, when considering overall hazard rates, aging is hidden from view by mortality associated with environmental factors, which predominates up to ages of 30-35 years in both genders.