In this study, a binary logit model was calibrated for mode choice decisions in freight transportation. Freight transportation system was undertaken within the context of logistics chain and analyzed by logistics cost model. Logistics regression analysis method of multivariate statistical analysis technique was used to calibrate the binary logit model. In the analysis, data related with the ceramics transported to the city of Antalya by trucks and combined transport (road and rail) were used. The model with the maximum forecast performance was chosen as the logit mode-choice model based on logistics costs. Sensitivity analysis carried out by using the chosen model have shown that the most important decision variables for mode choice in freight transportation are transportation cost, transportation time and transshipment delays. Transport policy proposals were made to promote the combined transport in the selected corridor.