We study the operational problem of loading and unloading trains at a container terminal. Trains are served by two gantry cranes which spread over multiple trains on parallel tracks at the terminal landside. Multiple straddle carriers are available to move containers between the stacking area and drop off locations perpendicular to the train tracks. The trains must be loaded and unloaded with given departure times. We develop a mixed-integer programming model to schedule the gantry cranes and straddle carriers to load and unload the trains. The objective is to minimize the total delay of trains. Due to the complexity involved, the problem is solved using a simulated annealing heuristic. We perform extensive numerical experiments to analyze the impact of different variables and parameters on minimizing the total delay. More specifically, the impact of three variables including the number of trains, containers on trains, and available straddle carriers are evaluated.