An investigation into the application potential and challenges of driving simulators in driving behavior research was conducted through a systematic literature review and bibliometric analysis. A total of 1 110 publications indexed in the Web of Science core collection and 240 articles from the CNKI core database were selected as the dataset and analyzed using CiteSpace for knowledge domain visualization. Three primary application domains of driving simulators in traffic research were identified and summarized: intelligent extraction of driving behavior characteristics, construction and validation of driving behavior models, and analysis of factors influencing driving safety. In the context of intelligent mining of driving behavior features, the interaction between drivers and advanced driver assistance systems (ADAS) or automated driving functions was examined under simulated conditions. With regard to safety-related behavioral analysis, influencing factors were categorized into three dimensions—driver state, traffic environment, and driver training—and systematically evaluated based on simulator-based experimental data. For model development and validation, representative computational models and associated experimental protocols were reviewed and discussed. Critical components of driving simulation experiments were further elaborated, including scenario design, data preprocessing, selection of behavioral indicators, and modeling procedures. Established methodologies for experimental design and data analysis in simulator studies were synthesized and presented. Future research directions were proposed, such as human-machine shared driving training, functional development of ADAS, simulation of driving behaviors in vehicle-infrastructure cooperative systems, and AI-driven modeling of driver behavior. It is indicated that driving simulators have been widely adopted in behavioral research. However, the validity and realism of experimental setups are still subject to limitations. With advancements in multi-modal data acquisition technologies and artificial intelligence, it is suggested that future studies should expand both the breadth and depth of simulation experiments to better capture the complexity and dynamic nature of real-world driving behaviors. Moreover, challenges remain in accurately replicating naturalistic traffic environments, enhancing data collection precision, and refining modeling techniques. These issues are expected to be addressed through interdisciplinary collaboration and integration of emerging technologies in the transportation and intelligent mobility domains.