Water-lubricated stern bearing (WSB) is a vital part of the ship propulsion-shaft system, and it is of great significance to monitor and analyze its lubrication status through film thickness data to improve the equipment operational reliability. In this paper, a full-size, large length-to-diameter ratio WSB experiment is carried out, and multi-sectional journal displacement data are collected under offset load. Accordingly, a bearing film-thickness identification model is established, which can identify the dynamic film thickness data in the circumferential direction of bearing by limited measurement points. On this basis, the film thickness distribution of the full bearing is obtained by combining finite element (FE) simulation and particle swarm optimization (PSO) algorithm. The effect of different speeds on the distributed lubrication characteristics of WSB under offset load was systematically analyzed based on film thickness data. Results show that the maximum identification error of the bearing film-thickness identification model is less than 7%. The bearing lubrication state changes dynamically as the speed increases, and the hydrodynamic lubrication effect in the middle of the bearing is enhanced. The area of each lubrication sub-region varies nonlinearly. Research results are instructive for further determine the service life of the shaft system.