Abstract In this paper, a new Simulation-Based Optimization Model (SBO-Model) is proposed to solve scheduling problem in stochastic multimodal freight transportation systems. The model is applied to find optimal services schedule in a real-world case study. In order to handle demand and travel time inherent variability, the stochastic service network design problem is addressed. Simulation modeling is used to efficiently account for real stochastic behavior with skewed continuous distributions. Such distinctive distribution shapes were commonly reported in transportation research studies that addressed the travel time reliability modeling. Results indicate that the SBO-Model can indeed provide reliable service schedules even under realistic complex stochasticity. The main finding is that, in order to solve efficiently such stochastic optimization problem, we need to go beyond the mean and variance estimates by considering the empirical distributions of uncertain parameters. Specifically, when the data exhibit skewness and/or multimodality, which are commonly found due to the traffic congestion. The originality of this work lies in the integration of stochastic models, commonly used in the transportation research field, for solving logistics planning problem generally addressed by Operations Research community.