Few would deny the contribution of vocational education on economic growth and social development, but the spatial dynamics behind the economic role of vocational education in transition economies has not been examined by the literature on economics of education. Specifically, two hypotheses are tested. First, the economic growth and vocational education development have significant global spatial autocorrelation, which means the development of economy and vocational education of one province depends on the economic or education level of neighboring provinces. Second, the economic growth and vocational education development have significant local spatial autocorrelation. With per capital GDP and vocational education scale data of 31 provinces in China from 1995 to 2008, both hypotheses are supported. Finally, the results show that the elasticity with spatial metrics is 1.522, which means the stronger economic role of vocational education because the elasticity is larger than 1, while the elasticity without spatial dynamics is only 0.926 which implies the weak economic role of vocational education. It also shows that the OLS model is confronted with the risk of spurious regression without considering spatial dynamics and the spatial error model is preferred because it’s robust.