Purposes: Among the growing population of the elderly in China, the percentage of empty-nest elderly increases over time. The empty-nest elderly lack companionship after their children's departure from home, and they often react with feelings of loneliness and depression. The incidence of depression in empty-nest elderly is significantly higher than that in non-empty-nest elderly. Studies have confirmed that social participation has a positive effect on mental health of the elderly, and it is necessary to further investigate the influence of social participation on the developmental trajectory of depression among the elderly, especially the empty-nest elderly. Based on longitudinal data, the current study compare the developmental trajectories of depression between empty-nest and non-empty-nest elderly and investigate the impact of social participation on their depression trajectories by applying a multi-group latent growth model. Method: The current study selected data from the Chinese Longitudinal Healthy Longevity Survey (CLHLS) from 2002 to 2014. Valid data from 1681 participants, including 799 males (47.53%), with an average age of 72.8 years (SD = 6.72), was analyzed. Participants were divided into empty-nest and non-empty-nest elderly based on one of the items from the survey. Then the data was fitted in a multi-group latent growth model to compare the developmental trajectories of depression between empty-nest and non-empty-nest elderly. Furthermore, social participation was incorporated into the model as a covariate in order to investigate its impact on the depression developmental trajectories of the empty-nest and non-empty-nest elderly. Results: (1) The overall trajectories of depression in the elderly followed a two-stage piecewise linear path, as revealed by a turning point on the trajectories. But there were differences between the trajectory of empty-nest and non-empty-nest elderly. The initial level (intercept) of depression of empty-nest elderly was higher than that of non-empty-nest elderly (empty-nest elderly: = 6.53, non-empty-nest elderly: = 6.28). At the first stage, the depression of non-empty-nest elderly, with a larger first-stage slope value, increased faster than that of empty-nest elderly (empty-nest elderly: = -.03, non-empty-nest elderly: = .09). At the second stage, the depression of empty-nest elderly increased faster than that of non-empty-nest elderly, as revealed by a larger second-stage slope value of depression trajectory of empty-nest elderly (empty-nest elderly: = .15, non-empty-nest elderly: = .06). (2) Social participation reduced the initial level of depression of the elderly (empty-nest elderly: = -.56, non-empty-nest elderly: = -.40). For empty-nest elderly, social participation could also slow down the growth rate of depression at both stages ( = .47, = -.16). For non-empty-nest elderly, social participation had no impact on the growth rate of depression. Conclusion: (1) There is a turning point on the depression development among empty-nest and non-empty-nest elderly. The depression of empty-nest elderly has no obvious change before the turning point, and it increases rapidly after the turning point. (2) The depression trajectories of empty-nest and non-empty-nest elderly are different. Empty-nest elderly are at high risk of depression, and have higher level of depression than non-empty-nest elderly. The growth rate of depression of empty-nest elderly is higher than that of non- empty-nest elderly at older ages. (3) Social participation is an important way to reduce and relieve depression among the elderly.