Late-life depression poses a significant concern in China, particularly among older rural women who experience higher prevalence rates and heightened risk. This dissertation investigates the factors influencing late-life depression in this specific high-risk demographic. The first essay identifies early life factors and adult physical health as prominent predictors of late-life depression among rural Chinese women using machine learning models. The second essay explores a relationship between education and depression among middle-aged and elderly rural Chinese women, revealing a negative education gradient in late-life depression. The third essay employs a Fuzzy Regression Discontinuity Design to assess the long-term impact of improved educational attainment on late-life depression, finding that increased schooling years lack an evident protective effect on late-life depression among the overall rural women population.