<p>Eco-driving behaviors among taxi drivers can significantly contribute to reducing urban air pollution, particularly in megacities such as Tehran. This study aimed to identify the psychosocial predictors of air pollution-reducing behaviors among taxi drivers based on the Health Belief Model (HBM). A cross-sectional descriptive-analytical study was conducted in late 2024 with 401 taxi drivers in Tehran, using structured face-to-face interviews. A validated questionnaire assessed seven HBM constructs and eco-driving behaviors. Pearson correlation and stepwise multiple regression analyses were applied. The findings revealed that self-efficacy was the strongest predictor of eco-driving behavior (β = 0.324, <i>p</i> &lt; 0.001), followed by awareness, perceived benefits, and perceived susceptibility. The final regression model explained 27.1% of the variance in eco-driving behavior (<i>R</i>² = 0.271). Although perceived barriers showed a significant negative correlation with eco-driving, it was not a significant predictor in the final model. These results highlight the importance of enhancing drivers’ self-efficacy and awareness of eco-driving benefits, while addressing motivational and contextual barriers. Developing targeted, theory-driven educational interventions based on the HBM may play a critical role in promoting eco-driving practices and improving air quality in high-pollution urban environments.</p>