Some times there may be many diÂ®erent ways to model a particular problem,rnchoosing the best one minimizes the complexity of the problem and time to solve.rnSince, as we have said earlier, any programming problem with constraint matrixrnstructure the same as the transportation problem is regarded as a transportationrnproblem regardless of it's physical meaning and because of it's simple structure,rnmodeling problems as transportation problems requires much less eÂ®ort to solvernthan that of modeling it diÂ®erently.rnIn this thesis the nonlinear transportation problem is considered as nonlinear pro-rngramming problem and algorithms to solve this particular problem are given. ThernÂ¯rst algorithm is similar to that of the transportation simplex algorithm exceptrnthe nonlinearity assumption. The second algorithm is dependent on the simplexrnalgorithm of Zangwill [32] that we modiÂ¯ed it to use the special property of therncoeÂ±cient matrix of the transportation problem so that we may take shortcuts tornmake problem solving simple. However the algorithms are not been compered tornother previous algorithms. Therefore in the future further work must be done :rn1. To measure the eÂ±ciency of the algorithm.rn2. To check how near the solution of the approximated problem of the piece-rnwise linear transportation problem is to the optimal solution of the originalrnproblem. And Â¯nallyrn3. to implement the algorithm to the real world problem.