Drawing on Reinforcing Spirals Model theory and longitudinal network analysis methods, we analyse the co-evolution of climate change communication network and online expressive participation around climate policy issues on China’s top social media platform Weibo (2019–2021). We find a mutual influence between actors’ engagement in communication relationships and their participatory behaviours over time. However, this dynamic does not operate at the individual level. Active/popular actors in the communication network do not necessarily become more participatory later, nor does higher participation significantly lead to greater activity/popularity in the subsequent time point. Instead, this dynamic is characterized by concurrent homogeneity-based network selection and network influence processes. Model results show that actors with similar participation levels are more likely to form future communication ties and actors with existing communication ties are more likely to converge in participation levels over time. This indicates a homogeneity-based reciprocal influence between network connections and individual engagement. We also find that endogenous network structural factors play a significant role in shaping people’s engagement with climate change communication. These findings underscore the importance of the social relational dimension in the dynamic interplay between communication and online expressive participation, contributing to a more nuanced understanding of the Reinforcing Spirals Model theory.