Timely and reliable truck activity data are of importance for transportation planning and investment analysis, traffic management, environmental and safety analyses, and operation and maintenance of infrastructure. Therefore, state and regional agencies have used various types of sensors to obtain truck flows. To measure accurate traffic flows with the use of sensors, considerable effort has been made to determine optimal sensor locations. The most common approach has been to focus on capturing the maximum number of origin–destinations (O-Ds) and routes at the sensor locations; this approach is referred to as the observability problem. However, when the long travel distances of trucks are considered, sensor locations that seek observability of O-Ds and routes may not acquire meaningful proportions of truck movements in a large-size truck network when budgets are limited. Therefore, this study provides a decision model that optimally locates sensors to capture the maximum O-D and route flow. A goal-programming approach with different weights that prioritize O-Ds or route flow, depending on a prior objective, was investigated. The proposed model was implemented in a real network in Los Angeles, California, with actual truck flow data obtained from sampled truck GPS trajectories. Results showed that significantly larger numbers of O-Ds and route flows were captured by the proposed flow-capturing model than with conventional observability approaches. In addition, trade-offs between O-D and route flow capturing depended on the applied weights and budgets. Particularly in the Los Angeles truck network, equally applied weight provided balanced capturing of O-D and route flow when sufficient numbers of sensors were located in the network. The proposed goal-programming approach can provide insights for researchers and practitioners for alternative sensor location solutions.