In this study, two parallel injection machine scheduling problem under crane constraint is considered. In setup phase of production with injection machine, the mold of the product is fixed to the machine. Molds are transported by a crane to relative machine. The crane is shared by two machines so setup times should not overlap. In addition, the problem has sequence dependent setup times and machine eligibility constraints. The objective is minimization of the makespan. So, class of the problem is P-2, S-1 vertical bar STsd, M-j vertical bar Cmax. A mathematical model was proposed for the problem and due to, the problem is in the class of NP-hard problems, a genetic algorithm is developed for solving large size problems. The success of the algorithm is shown by using randomly generated instances through mathematical model and lower bounds.