The field of systems biology, in dealing with the problem of reliability (“robustness”), incorporates theoretical and experimental studies of quantitative characteristics and mechanisms of failures and renewal processes in biological systems. It also includes elaboration of methods for testing reliability and predicting failures . Here, the results of application of the systems reliability approach to the problems of aging are presented. This approach is based on the simple general principles that (i) all biomolecular devices are designed in keeping with the genetic programs in order to perform the preset functions; (ii) all of them operate with the limited reliability, namely, for each and every bionanoreactor, normal operation acts alternate with accidental malfunctions; (iii) the preventive maintenance, i.e. – timely replacement or prophylaxis of unreliable functional elements, the metabolic turnover in cells and tissues, is the main line of assuring the high systems reliability; (iv) there is a finite number of critical elements which perform supervisory functions over the preventive maintenance; (v) the “supervisors” also operate with the limited, preset, reliability [2, 3]. On this reliability basis, the universal features of aging, such as the exponential growth of mortality rate with time and the correlation of longevity with the species-specific resting metabolism are naturally explained. The stochastic malfunctions in operation of the mitochondrial electron transport nanoreactors, that produce free superoxide radicals, seem to be of first importance. Basing on the reliability-theory approach, one can estimate that longevity of human brain could reach 250 years should the antioxidant defense against the free-radical failures be perfect. Thus, aging is the stochastic consequence of the limited (preset) reliability of biomolecular devices in cells and tissues. Furthermore, the systems reliability approach serves as heuristic methodology for preventive medicine including synthesis of novel radiation protectors, based on stable magnetic isotopes. [Supported by RFBR, project no. 14-04-00593a].