[Objective] Tourism functions not only serve as an important pathway for promoting the realization of national parks’ public welfare objectives, but also represent one of the key links in achieving the synergy between ecological conservation and sustainable development. From the perspective of online travel videos, this study explores the pathways through which perceptions of cultural ecosystem services (CES) in national parks impact potential tourists’ travel intention, thereby providing theoretical and empirical references for further refined management of national park ecosystems. [Methods] Taking China’s five officially established national parks as case studies, online travel videos of national park tourism from 2015 to 2024 were screened on mainstream video-sharing platforms, and typical interaction moments between video footage and viewers’ comments were systematically extracted. Furthermore, this study employed grounded theory to investigate potential tourists’ perceptions of CES in national parks, and, based on this, analyzed their travel intention using the fuzzy-set qualitative comparative analysis (fsQCA) method. [Results] (1) From the perspective of online travel videos, perceptions of CES in national parks involved five dimensions: sensory-aesthetic experience, recreational and entertainment experience, knowledge and educational experience, sense of place and cultural identity, and national park identity. (2) The potential travel intention expressed by video viewers mainly followed five pathways: sensory immersion-driven, entertainment interaction-driven, educational identity-driven, cultural identity-driven, and ecological responsibility-driven. (3) The potential travel intention expressed by viewers was essentially a psychological process jointly driven by cognitive construction and emotional experience. Through interactive participation, it triggered the viewers’ cognitive and emotional value identification with CES of specific national parks, thereby ultimately forming travel motivation. [Conclusion] Online travel videos can effectively overcome spatiotemporal limitations and extend the benefit boundaries of CES in national parks. Potential tourists’ travel intention follows a synergistic driving mechanism of emotional experience and cognitive construction. Future new media communication for national parks should strengthen the deep integration of emotional resonance and value identification, thereby promoting the realization of public welfare objectives.