We propose a new evolutionary computational model based on the Potts system. This hybrid approach combines a Genetic Algorithm (GA) with the principles of physical spin systems. The method, called Potts-based Genetic Algorithm (P-GA) involves the evolutionary search principles and those of the Potts's dynamics. Actually, the recombination operation between two individuals is performed by a Potts probability. Numerical experiments show that the proposed model performs similarly or sometimes even better than the standard GAsfor several difficult problems.