The Permutation Flow Shop scheduling Problem (PFSP) is a typical combinatorial optimization problem. In order to improve the efficacy in solving the PFSP, we applied a discrete mechanism to convert the real value of individuals into discrete job sequences at first. In particular, a Chaos-based Firefly Algorithm (CFA) is used to optimize the initial population, which provided a superior initial environment and improved the quality of optimization. In addition, the proposed method is tested by seven famous classic benchmark worksheets and compared with Particle Swarm Optimization (PSO) and a Genetic Algorithm (GA). At last, the simulation results show that the proposed CFA outperforms the others.