1 Transportation Research Record: Journal of the Transportation Research Board, No. 2663, 2017, pp. 1–11. http://dx.doi.org/10.3141/2663-01 Driving simulators have been widely used in transportation research and have potential applications for toll plaza safety research. The University of Puerto Rico at Mayagüez (UPRM) and the University of Massachusetts, Amherst (UMass-Amherst) performed a collaborative investigation using driving simulators to evaluate drivers’ behavior in two toll plazas with different signage and lane configurations that oper- ate under the U.S. jurisdiction. The studied toll roads were the Caguas South Toll Plaza in Puerto Rico and the West Springfield Toll Plaza in Massachusetts. The major safety issues identified in both toll roads were unexpected lane changes, sudden vehicle stops, and variable speed patterns. The purpose of this study was to exchange research scenarios between UPRM and UMass-Amherst to test drivers who were unfamiliar with the areas of the study and enlarge the scope. Assuming that the patterns of behavior were similar, the results would suggest that drivers’ behaviors from different regions depend largely on the geometry of the toll plaza and not on the driving culture particular to a region. This study will greatly add to the utility of driving simulator studies because the results reported from one region and one toll plaza arrangement should generalize to other regions around the country and to territories. Results show that familiar drivers had a better driving performance, with respect to variability of lane position, when compared with unfamiliar drivers. However, the proposed treatments for each toll plaza improved road safety for both familiar and unfamiliar drivers. The construction and operation of toll plazas has been in continual change because of the evolution of transportation technologies. Toll roads have been designed and used for more than five decades in the United States and Puerto Rico. However, there has been no real consistency across toll plazas in critical aspects of their design. Spe- cifically, traffic control devices and lane operations have varied con- siderably among operating agencies. Although the 2012 revision of the Manual on Uniform Traffic Control Devices (MUTCD) included a section dedicated to toll road signs (1), many toll plazas that were constructed before 2009 have not implemented these standardized signage configurations. Furthermore, emerging intelligent transpor- tation systems applications, such as electronic toll collection (ETC) and open toll road (OTR), have added to the complexity of the toll plaza environment and affected the safety of all road users. ETC lanes are automated toll collection systems that use wireless tech- nologies for the transactions of moving vehicles. OTR uses ETC systems while drivers are traveling at normal highway speeds (1). Although these payment methods do not require vehicles to com- pletely stop at the toll plaza, therefore having an important role in reducing vehicle emissions and other environmental aspects (2, 3), they raise the percentage of total crashes in main-line toll plazas associated with ETC lanes (4). The increase in vehicle crashes may be influenced by conflict points before the toll plaza station, including merging movements, queuing, acceleration and deceleration rates, speed variations among drivers, number of lanes, and signage configuration (4). Yang et al. indicated that crashes are more likely to occur in multilane toll plazas as a consequence of conflict points generated by drivers confused when deciding which toll lane to use (5). In addition, the variation in drivers’ speed in ETC lanes has increased the number of rear-end crashes, sideswipe crashes, and vehicle collisions with toll plazas as compared with the period before the implementation of the E-ZPass lanes, which is a type of ETC system (4). The sudden increase in the number of crashes that has occurred in toll plazas that operate multiple toll lane configurations has gener- ated the need to study driving behavior (6, 7 ). One of the research tools that can be used for toll plaza safety and human factors studies is driving simulators. Simulation studies are useful for the evaluation of emerging technologies and alternative solutions in transportation systems (8). For example, they have been used to study the effect of allowable permissive left-turn indications, effects of warning mes- sages and variable speed limits, effectiveness of yield markings in crosswalks, driving performance at high speeds, and the effect of alcohol, drugs, and various medical treatments on drivers’ perfor- mance (8–12). One of the major advantages that simulation provides is the ability to expose participants to potential hazardous situations without physically harming the subject driver, making it possible to evaluate dangerous scenarios with a high collision risk. Transportation studies have been undertaken using driving simula- tors to understand the causes of crashes at toll plazas and to evaluate alternative treatments. For example, Valdés et al. found that over- head signing configuration had a better potential for crash reduction than roadside signage when providing information to drivers indicat- ing the appropriate lane use and posted speed limits (13, 14). They did so by using a driving simulator to evaluate Puerto Rican drivers Comparative Analysis of Toll Plaza Safety Features in Puerto Rico and Massachusetts with a Driving Simulator Didier Valdés, Benjamín Colucci, Michael Knodler, Donald Fisher, Bryan Ruiz, Johnathan Ruiz, Ricardo García, Enid Colón, and Foroogh Hajiseyedjavadi D. Valdés, B. Colucci, B. Ruiz, J. Ruiz, R. Garcia, and E. Colón, Department of Civil Engineering and Surveying, University of Puerto Rico at Mayagüez, P.O. Box 9000, Mayagüez, Puerto Rico 00680. M. Knodler, D. Fisher, and F. Hajiseyedjavadi, Department of Mechanical and Industrial Engineering, University of Massachusetts, Amherst, 142B Marston Hall, Amherst, MA 01003. Corresponding author: D. Valdés, didier.valdes@upr.edu.