In this study, tour-based travel demand models were developed to describe the travel pattern of courier vehicles; the models overcome the limitations of four-step freight demand modeling. This study used a microsimulation-based modeling framework. The study area, Seoul, South Korea, was divided into block-based smaller traffic analysis zones, and the travel data from real-world courier service companies were used for model development and validation. The developed tour-based urban freight demand models were composed of eight steps: tour start, departure time choice, next-stop destination choice, vehicle movement, stop duration, next-stop purpose choice, return, and tour termination. After specific models were developed for each of the eight modules, the proposed modeling framework was applied, and the results were compared with the data observed in regard to average trip distance, trip length distribution, accuracy, sensitivity, specificity, number of stops, and average travel distance of the tours. Overall, results of the proposed models were reasonable from the perspective of urban freight demand modeling.