The aim of this study is to assess, whether estimation of the same innovation-adoption model at farm- and municipality-level results in an ecological fallacy, meaning that based on aggregated data, one would make inverse inferences about the driving forces influencing the adoption decision at the farm level. The adoption of an emerging alternative crop in Austria, the Styrian Oil Pumpkin, serves as an applied example. Our findings indicate the presence of an ecological fallacy. We therefore propose further research, which could consist of Monte Carlo simulations in order to analyse sensitivity of results with respect to the degree of aggregation.