Multi-level programming problem (MLPP) is an optimization problem whichrnhas other optimization problems in its constraint set and has a decision makerrnfor each objective function controlling part of the variables. In this thesisrnwe try to investigate some special multilevel programming problem whichrnis a tri-level programming with the quadratic fractional objective functionsrnand polyhedral constraints. Fuzzy goal programming approach is one of thernmethods used for solving multilevel programming problems through fuzzyrngoal programming (FGP) model formulation, corresponding objectives ofrnequivalent multi-level programming problem are transformed into fuzzy goalsrn(membership functions) by means of assigning an aspiration level to each ofrnthem and suitable membership function is de_ned for each objectives. Thenrnachievement of the highest membership value of each of the fuzzy goals is formulatedrnby minimizing the sum of negative deviational variables. However,rndue to the conicting nature of each DM's objective function and hierarchicalrnnature of the problem, solving the formulated problem for one level may notrnproduce a satisfactory solution for the system. So this nature of the problemrncan be controlled by de_ning a new membership function of _rst level decisionrnmaker (FLDM) and second level decision maker (SLDM) based on thernsolution we obtained in the _rst formulation. Therefore, after some iterationsrnFLDM and SLDM arrive at a satisfactory level and this intern producerna satisfactory solution for the system.rnKeywords: Tri-level programming, Quadratic fractional programming, Fuzzyrngoal programming, membership functions, deviational variables