The paper addresses a multi-item, multi-plant lot-sizing problem with transfer costs and capacity constraints. The problem is reformulated according to a multi-commodity flow formalism, and decomposed, through Lagrangean relaxation, into a master facility location problem and a slave minimal cost multi-commodity flow problem. The decomposition framework gives rise in a natural way to the design of Lagrangean based heuristic. Numerical experiments showing the efficiency of the proposed approach are reported.