Choice Modelling applications can be designed to estimate main effects only or multiple-way interactions between attributes. It has been reported that higher order effects generally account for less than 10% of the choice explanation. Nevertheless, the amount of applications testing for interactions among attributes in environmental valuation is very limited. This paper reports a Choice Modelling exercise valuing climate change impacts on plant cover, land erosion and fire risk in Spanish shrublands. Two out of three two-way interactions were found significant and to account for more than 20% of the choice model explanation. Their contribution to the log-likelihood value was comparable to the one of the main effects variables. Moreover, accounting for second order interactions significantly altered the estimates of the implicit prices of attributes compared to the main effects specifications.