We　investigate　optimization　performance　of　chaotic　evolution　algorithm　with　random　crossover　rate　by　comparing　the　ones　with　constant　crossover　rate.　We　found　that　searching　around　the　target　vector　is　always　useful　for　optimization　by　analysing　convergent　results.　Setting　a　big　crossover　rate　at　the　initial　generation　and　small　one　in　the　end　are　great　strategies　of　crossover　rate　setting　for　chaotic　evolution.　It　presents　exploitation　and　exploration　capabilities　of　chaotic　evolution　algorithm.　However,　in　the　more　benchmark　cases＇　condition,　chaotic　evolution　algorithms　with　any　crossover　rate　setting　have　the　same　search　performance　from　multivariate　statistical　test.　We　found　that　search　with　a　large　crossover　rate　in　the　beginning,　and　do　with　a　small　one　at　the　end,　it　can　enhance　search　capability　of　chaotic　evolution　algorithm.　We　also　discuss　and　analyse　related　issues,　open　topics,　and　future　works　on　crossover　rate　of　chaotic　evolution　algorithm.