Portfolio selection problem is generally a nonlinear programming which has been solved by a variety of heuristic and non-heuristic techniques. This paper presents a novel heuristic method for solving an extended fuzzy portfolio selection model. The significance of this paper is twofold. First, it extends fuzzy mean-variance-skewness model to fuzzy mean-variance-skewness-kurtosis model. Second, a powerful heuristic called Firefly Algorithm (FA) is proposed for solving model. No study has ever proposed and solved this expanded model. Finally, several numerical examples are provided to illustrate the modeling idea and performance and effectiveness of the proposed algorithm is compared against the exact approach (LINGO software) in terms of fitness value and required computational time. Results show that the proposed FA is very promising and achieves quality results for fuzzy portfolio selection in a reasonable time.