This paper constructs an environmental impact model of Fine Particulate Matter (PM2.5) concentration using the STIRPAT analysis framework and empirically tests the impact of waterway transport on PM2.5 concentration and its spatiotemporal heterogeneity based on provincial panel data in China from 1998 to 2019 using the Geographically and Temporally Weighted Regression (GTWR) model. The empirical results show: (1) The GTWR model, which considers spatiotemporal factors, outperforms the global regression (OLS) model in terms of fitting effects; (2) The impact of waterway transport on PM2.5 concentration exhibits both positive and negative effects. Specifically, the positive effect of waterway goods transport on PM2.5 concentration is significant, while the negative effect of waterway passenger transport is significant, and the two present an approximate substitution pattern in spatial distribution; (3) The effects of various influencing variables on PM2.5 concentration all exhibit positive and negative effects as well as varying degrees of fluctuation, and they display significant spatiotemporal heterogeneity. From a temporal perspective, different influencing factors show certain evolutionary trends, while from a spatial perspective, the differences in influencing factors between regions are significant. These findings not only provide empirical support for regionally differentiated PM2.5 control strategies in China but also offer valuable insights for other developing countries facing trade-offs between transportation development and air quality. The study contributes to the global discussion on sustainable transportation planning and environmental governance by highlighting the nuanced roles of freight and passenger transport in pollution outcomes. Based on the empirical conclusions, this paper proposes related policy recommendations for controlling PM2.5 emissions.