The rapid growth in e-commerce activities and the constant specialization of industries have aroused an unparalleled demand for trucking in urban areas, leading to growing concern over its interference to the transportation system. Understanding truck route choice preferences across long-haul, medium-haul, and short-haul trips can offer insights for designing the truck route network tailored to specific freight demand types, so as to effectively reduce their interference to passenger transportation. However, limited research has been conducted to explore the heterogeneity or similarity of route choice preferences across those freight demand types. This study utilizes the Path Size Logit Model to explore the characteristics of preferred route across long-haul, medium-haul, and short-haul trips, and reveal the underlying route choice mechanism behind enormous trucking activities. By employing truck GPS data from China, we find that (1) although the characteristics of preferred routes vary across long-haul, medium-haul, and short-haul trips, those trips collectively reflect full consideration of travel efficiency, safety, and reliability; (2) all these freight demand types incline to the routes with shortest travel distances instead of those with shortest travel time, while short-haul trips exhibit the highest sensitivity to travel distance; (3) drivers in both long-haul and medium-haul trips favor routes that combine motorways and sub-arterial roads, while long-haul trips present higher sensitivity; (4) drivers in short-haul trips show preferences for routes featuring fewer turns, and sub-arterial roads given last-mile delivery demand. Finally, we propose suggestions for designing urban truck route network to accommodate diverse freight demand in high-density urban areas with limited road resources.