We present a two-phase heuristic approach for the Hybrid Flexible Flowshop with Transportation Times (HFFTT) which combines a metaheuristic with constraint programming (CP). In the first phase an adapted version of a state-of-the-art metaheuristic for the Hybrid Flowshop generates an initial solution. In the second phase, a CP approach reoptimizes the solution with respect to the last stages. Although this research is still in progress, the initial computational results are very promising. In fact, we show that the proposed hybrid approach outperforms both the adapted version of and earlier CP approaches.