This paper empirically examines the causal relationship between energy consumption and economic growth, by controlling with CO2 emission and economic globalization in multivariate models, in a sample of the 91 less developed, developing and developed countries individually, in the period of 1970-2013. I apply the Toda-Yamamoto augmented Granger non-causality testing procedure. I observe a significant unidirectional causality running from energy consumption to economic growth in twenty-one countries. In thirty-one countries, conservation hypothesis is valid. Empirical evidence also shows that there is a feedback hypothesis in sixteen countries and no causal relation in only twenty-three countries.