In this paper, a two-stage hybrid flowshop scheduling problem with independent setup times is examined. The complexity of the studied problem is NP-Hard in the strong sense since it generalized the two-stage hybrid flowshop scheduling problem. Several real-life problems in different areas are modelled using the addressed problem, as in parallel computing and manufacturing processes. Solving the current scheduling problem necessitates the construction of adequate algorithms, providing nearoptimal solutions within a satisfactory computing time. In this study, a genetic algorithm with specific features and three other heuristics are developed. These three heuristics are based on the optimal solution of the parallel machine scheduling problem with release dates and delivery times. In order to assess the performance of the meta-heuristic and the heuristics, a family of three lower bounds is proposed. An exhaustive numerical study is performed over a total of 1920 test problems. The obtained results provide strong evidence of the efficiency and the performance of the proposed procedures.