This research focuses on nonlinear minimum spanning tree problems with linear constraints. A approximate solution method through tabu search is proposed. The results of numerical experiments show some advantages of the proposed tabu search method in comparison with genetic algorithms.