This paper aims to measure carbon market risk considering heterogeneous dependence in multiple risk factors including carbon price, domestic interest rate and exchange rate. Based on vine copula, we analyze dependency and then explore the risk, and we study periodic differences across three stages in the EU ETS. We find vine copula considers risk to a greater extent by heterogeneous dependence in risk factors, and such consideration decreases excessive risk aversion. Moreover, both risk factor dependence and VaR have significant periodic differences. This study provides a more accurate and reasonable risk measurement method to better realize carbon reduction targets.