Fuzzy multi-objective zero-one linear programming (FMOZOLP) has many applications in various fields such as assembly line balancing, assignment, project portfolio selection and maximal covering location problems. In many of the existing methods for solving FMOZOLP problems, membership degree of different points of a fuzzy number is not considered or by performing α-cut, points with membership function more than or equal to α, are included in calculations. However, even in this case, membership degree of these points has no effect on optimal solution. In this paper, in addition to modifying defects and failures of Yu and Li method  in solving fuzzy zero-one linear programming problems, we develop a novel approach to solve FMOZOLP problems considering membership degree of coefficients. Finally, an illustrative example for the project portfolio selection is included to compare results obtained by the proposed approach with results obtained by the other fuzzy methods.