More than 40 years have passed since the original publication by Hayflick and Moorhead led to the concept of the 'Hayflick limit' of the maximum number of divisions which somatic cells undergo in vitro. This concept is still regarded as a fundamental characteristic of species longevity. Here we want to emphasize another characteristic of somatic cells, namely, the duration of their survival in vitro in the non-dividing state after cessation of proliferation. This is suggested on the basis of results of recent experiments with so-called Japanese accelerated senescent mice. Results of these experiments reveal a good correlation between the longevity of the mice, the number of duplications of their fibroblasts in vitro, and the survival time of these cells in the non-dividing state. In routine culture conditions, cell survival time may be very long, as much as a few years. However, when the cells are grown under conditions of oxidative stress, cellular longevity is markedly shortened. This new test may serve as an additional marker of organismic longevity. The comparative value of both tests, the classical 'Hayflick limit' and the new test, is discussed.