Over the past few decades, the proportion of elderly people with depression in Korea has been rapidly increasing. The purpose of this study is to explore relative effects of health and family factors on geriatric depression. The data sources are from Korean Longitudinal Study of Aging(2008). The Analysis sample consists of 4,040 cases. Analysis results show that there is a significant variability in geriatric depression according to sociodemographic characteristics, health characteristics, and family characteristics. Hierarchical logistic regression analyses are employed to assess the causal relationship between health characteristics and geriatric depression. Model II showed that ADL(exp(B)=1.732), having difficulties due to sight deterioration(exp(B)=1.398), having difficulties due to masticatory force deterioration(exp(B)=1.414), having difficulties due to pain(exp(B)=2.435), and subjective health status(exp(B)=2.010) are reported as predictors of geriatric depression. Namely, the probability of having geriatric depression of the above predictors has been 1.7 times, 1.4 times, 1.4 times, 2.4 times. 2.0 times higher than normal figures. Among health characteristics, pain is the most prevalent factor. Model III showed that as for family characteristics, coresiding with children(exp(B)=1.312) and frequency of contacting with children by phones, letters, or emails(exp(B)=1.477) are reported as predictors of geriatric depression. Key findings are that health factors have larger effects than family factors on geriatric depression.