In many fields of the economics discipline, much of the empirical work includes a thorough analysis of time series data. In environmental economics, however, such an analysis is often neglected. This is unfortunate for two reasons. First, as Lee and List (2004). Examining trends of criteria air pollutants: ere the effects of government intervention transitory? Environmental and Resource Economics 29 (1), 21-37 argue, time series analysis can provide many new insights relevant in modelling work or in forwarding policy advice. Secondly, the nature of the time series has a profound impact on the modelling work. This paper shows that such an analysis is a necessity. We illustrate this with a Monte Carlo investigation of an Environmental Kuznets type of transition between non-stationary variables. The Environmental Kuznets Curve hypothesis posits an inverse U-shaped relation between environmental pollution and income. Although both pollution and income may be stochastically trending, the empirical literature has largely ignored this property. Through Monte Carlo experiments we show that with stochastically trending series, regression analysis spuriously confirms the EKC hypothesis in 40% of the cases.