<p>Urban greenways are increasingly promoted as key infrastructures for active mobility. However, the perceptual dimensions that shape cycling behavior along these corridors remain insufficiently understood. While prior research has predominantly focused on built environment attributes, the role of soundscape as an experiential and behavioral driver has received limited empirical attention. This study integrates crowdsourced trajectory data, street-level imagery, and machine learning techniques to examine the relationship between multidimensional soundscape features and cycling preferences along urban greenways, employing interpretable models to uncover underlying nonlinear mechanisms. The results indicate that soundscape attributes such as sound quality, eventfulness, and traffic noise exert significant yet nonlinear influences on cycling volume, suggesting that the relationship between acoustic conditions and cycling activity is more complex than a simple linear reduction in noise. By conceptualizing soundscape as an actionable perceptual resource rather than merely a noise-control issue, this study advances urban greenway research from physical performance evaluation toward experiential mobility planning. The findings provide quantitative evidence for integrating soundscape-oriented interventions into bicycle-friendly urban design, offering a cost-effective approach for promoting sustainable active travel.</p> Graphical Abstract <p></p>