This article uses econometric techniques to examine the effect of corporate carbon performance on corporate financial performance. I extend the existing literature in this research field by differentiating between two measurement perspectives: carbon performance expressed as annually reported carbon dioxide (CO2) emission equivalents and improvements in carbon performance over time. Thereby, the article re‐addresses the research question ‘when and how does it pay to be green?’ in the context of carbon emissions and climate change mitigation. Using a nonlinear modeling technique, the findings indicate that it pays to be green for companies with superior carbon performance but not for companies with inferior carbon performance. The results also show that carbon emission mitigation is linearly and significantly positive related to return on sales (ROS) but negatively related to Tobin's q. These contradictory findings help us to understand why – in spite of growing regulatory pressure – companies have been slow to respond with effective action to tackle climate change beyond marginal efficiency improvements that correspond to ‘low‐hanging fruits’. The empirical analysis is based on an unbalanced sample of 7625 firm‐year observations covering carbon emission data (Scope 1 and Scope 2) for 1640 international firms from 2003 to 2015. Copyright © 2017 John Wiley & Sons, Ltd and ERP Environment