To tackle the multiple difficulties of complex optimization problems in terms of high dimension, large scale, mixed-type variables, strong constraints, multiple minima, multiple objectives, dynamic and stochastic environments, etc, co-evolution is an effective way to improve the performances of evolutionary algorithms. An overview on co-evolutionary algorithms is presented in terms of population-collaboration, individual-collaboration, algorithm-collaboration, operatorcollaboration, parameter-collaboration, strategy-collaboration, and human-machine-collaboration. The mechanisms and designs of co-evolutionary algorithms are summarized. The applications of co-evolutionary algorithms in various fields are introduced. Finally, some future research direction and contents are pointed out.