In this paper, we study a mixed pallet collection problem in a warehouse of the company operating in fast moving consumer goods industry and present a mixed integer programming formulation with the objective function of total travelling distance minimization. The problem studied is shown to be equivalent to the well-known vehicle routing problem. Since the problem belongs to the class of NP-hard problems, introduced mathematical formulation cannot provide optimal solution in an acceptable amount of time. We, therefore, develop an algorithm based on Simulated Annealing (SA) meta-heuristic approach to find near-optimal solution in a quite shorter computational time. Routes are constructed using Clarke&Wright saving algorithm and then these routes are perturbed whereby three neighborhood operators, namely swap, insert, swap-range are utilized to further improve the quality of the solution. Experimental results based on a real case instance demonstrates that SA algorithm is capable of providing solution more quickly than that of CPLEX solver but the quality of the solution found by SA is 7% worse than that of CPLEX.