The lack of social support in elderly populations incurs real societal costs and can lead to their poor health. The aim of this study is to investigate the self-rated health (SRH) and social support among older people as well as its associated factors.We conducted a cross-sectional study among 312 urban community-dwelling elderly aged 65 to 90 years in Tainan Taiwan and Fuzhou Fujian Province from March 2012 to October 2012. A Spearson correlation test, independent t test, a Pearson χ test, a linear regression analysis, and a multiple-level model were performed to analyze the results.The participants identified children as the most important source of objective and subjective support, followed by spouse and relatives. Tainan's elderly received more daily life assistance and emotional support, showed stronger awareness of the need to seek help, and maintained a higher frequency of social interactions compared with the elderly in Fuzhou. The mean objective support, subjective support, and support utilization scores as well as the overall social support among Tainan's elderly were significantly high compared with the scores among Fuzhou's elderly. Further, Tainan's elderly rated better SRH than Fuzhou's elderly. Correlation analysis showed that social support was significantly correlated with city, age, living conditions, marital status, and SRH. Multiple linear regression analysis, with social support as a dependent variable, retained the following independent predictors in the final regression model: city (4.792, 95% confidence interval [CI]: 3.068-6.516, P = 0.000), age (-0.805, 95% CI: -1.394 to -0.135, P = 0.013), marital status (-1.260, 95% CI: -1.891 to -0.629, P = 0.000), living conditions (4.069, 95% CI: 3.022-5.116, P = 0.000), and SRH -1.941, 95% CI: -3.194 to -0.688, P = 0.003). The multiple-level model showed that city would impact older people's social support (χ = 5.103, P < 0.001). Marital status (-2.133, 95% CI: -2.768 to -1.499, P = 0.000), education (1.697, 95% CI: 0.589-2.805 P = 0.003), living conditions (4.20, 95% CI: 1.762-6.638, P = 0.000), and SRH (-3.144, 95% CI: -4.502 to -1.727, P = 0.000) were the associated factors. Thus, city, age, marital status, education, living conditions, and SRH might be the associated factors for social support among older people.This study presents some feasible implications for social support improvement in China and in other nations worldwide.