Purpose: The Purpose of this study is to build up model of the emotional labor worker's health. Methods: Data was collected from 230 emotional labor workers from 2 department stores, 2 insurance companies, and 3 hotels located in Seoul and Kyung-gi areas, by the health managers and team leaders. For data analysis, descriptive statistics, pearson correlation, factor analysis and covariance structure analysis were used by SPSS window 10 version and AMOS 4.01. Results: The hypothetical model showed a goodness-of-fit to the empirical data(GFI=0.90, RMR=0.04, NFI= 0.79, PNFI=0.64). Eight out of fifteen paths of the model were accepted, while the other seven paths rejected. : From 'surface acting' to 'health', from 'social support' to 'surface acting', from 'social support' to 'health', from 'factors of organizational culture' to 'surface acting', from 'factors of organizational culture' to 'deep acting', from 'personal factor' to 'social support', from 'factors of organizational culture' to 'social support', from 'routine stress' to 'social support'. In conclusion, it has been confirmed that surface acting and social support were dominating factors to the health of emotional labor workers. Conclusion: In order to manage the health of emotional labor workers, it is imperative to develop a strategy to reduce their surface acting.