The literature on the employment impact of China’s GVCs participation has focused on the substitution effects of Chinese inputs imported by developed countries, while few studies have been made on its domestic job creations. To complete this gap, this study proposes a GVCs augmented labor demand function, which is applied to panel data of 16 Chinese manufacturing industries over the 2005-2014 period using Arellano and Bond’s GMM estimator for dynamic panel data model specifications to estimate employment effects of different GVCs participation modes. The obtained positive coefficients of backward linkages show that increasing labor intensity processing and assembly increased employment, while the negative coefficients of China’s forward linkages and position mean that producing more capital intensity intermediate inputs to be embodied in third countries’ exports and upgrading along to GVCs provided less job opportunities. Thus, the decrease of processing and assembly exports, the growth of intermediated inputs’ exports and the position improvement during the studied period have all diminished the manufacturing employment. These results support the ‘mixed-blessing hypothesis’ of GVC participation in the literature.