The rapid adoption of electric vehicles (EVs) for inter-city travel has substantially increased the en-route charging demand along highways. Understanding EV drivers’ charging choice behavior can help managers manage charging facilities and improve the service quality. This study conducts a stated preference survey to obtain the charging choice decision data, which includes socio-demographic attributes, scenario attributes, and latent attributes. Then, a hybrid choice model (HCM) framework is applied to model the en-route charging choice behavior. The results show that the HCM latent class logit (LCL) model provides a good fit and reveals the heterogeneity among EV drivers. Two classes of EV drivers are identified in the HCM LCL model. The first class is service-oriented drivers, who value station service quality and display risk-averse, habitual behaviors. The second class is practical-oriented drivers, who prioritize cost, state of charge, and station accessibility. The findings highlight the importance of considering both observable and latent variables in modeling charging choice behavior. They also offer implications for differentiated station management strategies, such as enhancing service quality, applying dynamic pricing, and optimizing station spacing to balance charging demand across the network and improve user satisfaction.