This research aimed to develop a model to forecast container transport mode choice for processed agricultural products including tapioca starch, rice, and sugar from northeastern Thailand. The study applied the stated preference survey technique and developed a binary logit model from the results. It was found that the factors influencing container transport mode choice include transport time, cost, punctuality, availability of scheduling staff, and distance from the factory to railway station. The stated preference questionnaires were distributed, and responses were obtained from 19 manufacturers in the study area. The binary logit model was developed and proved to fit the real dataset. It was able to provide forecast precision to a satisfactory level. The calibrated model was tested with various transport policies to demonstrate possible approaches to improve rail transport mode share and promote a more sustainable freight transportation option.