Abstract: Gas-liquid annular flow is a widely applied flow pattern in engineering due to its safety and cost-effectiveness. It aims to investigate the characteristics and transition mechanisms of annular flow. The thin liquid film method (TLFM) was employed and the Taylor analogy breakup (TAB) model was utilized to simulate droplet breakup. The O’Rourke’s stochastic estimation algorithm was used to account for droplet coalescence, and the Rosin-Rammler distribution was used to optimize the size distribution of entrained droplets. This comprehensive approach replicates the dynamic equilibrium processes of entrainment, breakup, coalescence, and deposition within annular flow. To validate our improved TLFM, we conducted numerical simulations of classic annular flow experiments from the literature. The resulting droplet size distributions closely matched experimental data, confirming the accuracy of our model. We compared two key parameters of annular flow—entrainment fraction and liquid film thickness. The Mean Absolute Percentage Error (MAPE) was within 10%, demonstrating enhanced precision of the original TLFM. Finally, the model was used to simulate annular flow in pipelines under different superficial velocities, analyzing the impact of velocity on droplet distribution.