The empirical environmental Kuznets curve (EKC) literature is vast but far from conclusive. Many authors have analysed the existence of an EKC for various pollutants. Others have used the EKC framework to identify country characteristics that help to explain the income–environment relationship. In this framework environmental degradation is analysed using a second or third order polynomial in income and a limited number of control variables. Some authors question whether this standard framework is appropriate. This paper proposes an alternative to study the specific characteristics of countries that have experienced economic growth and an improving environment at the same time. We estimate a binary response model and find an EKC-like relation between the probability that a country's environment improves with economic growth and per capita GDP. Our evidence further suggests that the level of environmental damage is an important explanatory variable. We also confirm the importance of an open political system. These results indicate that the binary response model could be a valuable alternative to test which country specific characteristics are associated with a negative IER.