Utilizing Chengdu’s travelogue data from 2018, 2021, and 2023, this research applies social network analysis (SNA) and spatial stats to build annual tourism flow networks for Chengdu and its Central District. Objectives include understanding temporal evolution in Chengdu’s tourist flow network and pinpointing influencing factors. Findings reveal: 1) Chengdu nodes exhibit stronger functional capabilities than those in the Central District, with an average functional index of 1.72 times higher. Both scales’ node aggregation, radiation, and intermediary effects demonstrate instability. 2) The Central District boasts a higher network density (0.084) and stronger ties than Chengdu’s overall network density of 0.052. 3) The core drivers of Chengdu’s tourist flow network structure are attraction count (q = 0.91), tourist receipts (q = 0.74), and highway mileage (q = 0.66). Despite these insights, existing studies have certain limitations. First, the temporal resolution of the data (limited to three years) may not fully capture seasonal or short-term variations in tourist flow dynamics. Second, the current modeling approach does not account for dynamic factors such as policy changes or shifting tourist preferences, which could significantly influence network evolution. Future research could address these gaps by incorporating higher-frequency data and developing more dynamic models to understand the complexities of tourism flow networks better. These results provide valuable quantitative evidence to guide the premium development of Chengdu’s tourism sector and inform strategic planning for sustainable growth.