Gene expression programming (GEP) has been widely used in the areas of pattern recognition and knowledge discovery, however, when dealing with complicated problems, it is very time-consuming and the number of generations is large. In order to overcome these drawbacks, this paper proposes a multi-chromosomes GEP algorithm (MC-GEP). Firstly, the individual is composed of multiple chromosomes, each chromosome consists of one or more genes. Secondly, the expression of each chromosome or combinations of several chromosomes may be chosen to indicate the individual. Finally, chromosome recombination is changed and performed orderly like meiosis. Experimental results show that MC-GEP can reduce the running time and the number of generations with respect to the GEP.