A major criticism to the existing energy-growth-environment literature, we notice, is the selection of methodology. Panel estimation techniques that fail to consider both heterogeneity and cross-sectional dependence across countries may cause forecasting errors. The other concern related to the literature is that only a small number of studies analyze the influence of tourism on CO2 emissions even though tourism sector has potential for affecting the environment. To fulfill the mentioned gaps in the literature, this study analyzes the relationship among carbon emissions, real income, energy consumption and tourism for a panel of the EU and candidate countries over the period 1995-2011 by using heterogeneous panel estimation techniques with cross-sectional dependence. Results from the CADF and the CIPS panel unit root tests show that the analyzed variables become stationary at their first-differences. The LM bootstrap panel cointegration test indicates the presence of a long run relationship among the analyzed variables. Results from the OLS with fixed effects, the FMOLS, the DOLS and the group-mean estimator reveal that energy consumption contributes to the level of emissions while real income and tourism mitigate CO2 emissions. The Emirmahmutoglu-Kose panel Granger causality test suggests that there is one-way causality running from tourism to carbon emissions, and two-way causality between CO2 emissions and energy consumption, and between real income and CO2 emissions. Policy implications are further discussed.