This paper addresses a new version of the well-known vehicle routing problem with cross docking; we consider a more generalized version in which a customer orders different products from several suppliers. To solve this problem, we adapt a recently proposed matheuristic based iterated local search. In particular, we focus on the feasibility tests for insertions and compare the new developed methods with those already existing. Finally, computational experiments on instances adapted from the classical vehicle routing problem with cross-docking are reported. Computational experiments with problems with 30 to 200 nodes indicate that the proposed algorithms are very competitive compared to previously proposed approaches like Memetic algorithm and tabu search.