After successful application of fuzzy sets theory in many disciplines such as Management, Economic, Marketing and etc., this theory has been developed in the various subject as well as linear programming, fuzzy linear programming (FLP) have attracted many interests and various kinds of FLP is appeared in the literature recently. Semi-fully fuzzy linear programming problems have been investigated by many researchers in the recent decay. There are several methods to solve mentioned problems, but almost all of them need to have the basic condition for the initial solution. Hence, in this paper, we introduce a method on the primal-dual simplex algorithm which is works more efficiently than other methods by eliminating of the basic condition for the initial solution. A numerical example is given to illustrate the mentioned approach
