This paper applies panel vector autoregression (PVAR) along with a system-generalized method of moment (System-GMM) to examine the dynamic causal relationship between economic growth, carbon emissions and energy consumption for 116 countries over the period 1990–2014. Using multivariate model, the empirical results from this study have established key relationships which have important policy implications. First, at the global and regional levels, economic growth does not cause energy consumption. Second, with the exception of the global and Caribbean-Latin America, economic growth has no causal impact on carbon emissions, however, economic growth has a negative impact on carbon emissions at the global level and the Caribbean-Latin America. Third, carbon emissions positively cause economic growth. Fourth, energy consumption positively causes economic growth in sub-Saharan Africa while it negatively causes economic growth at the global, Middle East and North Africa (MENA), Asia-Pacific and Caribbean-Latin America. Fifth, energy consumption positively causes carbon emissions in MENA but causes carbon emissions negatively in sub-Saharan Africa and Caribbean-Latin America. Lastly, with the exception of MENA and the global sample, carbon emissions do not cause energy consumption. The impulse response function shows evidence of Environmental Kuznets curve at the global scale and sub-Saharan Africa. The policy implications of this paper are discussed.