Roadway tolls are designed to raise revenue to fund transportation investments and manage travel demand and as such may affect transportation system performance and route choice. Yet, limited research has quantified the impact of tolling on truck speed and route choice because of the lack of truck-specific movement data. Most existing tolling impact studies rely on surveys in which drivers are given several alternative routes and their performance characteristics and asked to estimate route choices. The limitations of such an approach are that the results may not reflect actual truck route choices and the surveys are costly to collect. The research described in this paper used truck GPS data to observe empirical responses to tolling, following the implementation of a toll on the State Route 520 (SR-520) bridge in Seattle, Washington. Truck GPS data were used to evaluate route choice and travel speed along SR-520 and the alternate toll-free Route I-90. It was found that truck travel speed on SR-520 improved after tolling, although travel speed on the alternative toll-free Route I-90 decreased during the peak period. A set of logit models was developed to determine the influential factors in truck routing. The results indicated that travel time, travel time reliability, and toll rate were all influential factors during peak and off-peak periods. The values of truck travel time during various time periods were estimated, and it was found that the values varied with the definition of peak and off-peak periods.