Abstract This paper considers the mixed-integer non-linear optimization of reliability–redundancy allocation problem (RRAP) to determine simultaneous reliability and redundancy level of components. In the RRAP, it is necessary to create a trade-off between component reliabilities and the number of redundant components with the aim of maximizing system reliability through component reliability choices and component redundancy levels. RRAPs have been generally formulated by considering an active redundancy strategy. A large number of solution methods have been developed to deal with these problems. In this paper, a cold-standby strategy for redundant components is used, for the first time, to model the RRAP; a modified genetic algorithm is developed to solve the proposed non-linear mixed-integer problem; and three famous benchmark problems are used for comparison. The results indicate that the cold-standby strategy exhibits a better performance and yields higher reliability values compared to the previous studies.