The Ricardian literature has only a handful of contributions addressing the presence of spatial heterogeneity in the marginal effects of climate change on agriculture. Although the majority of these studies offer models with group-specific slope parameters to account for spatial heterogeneity, large discrepancies on which grouping should be preferred still exist. This paper evaluates the extent to which expected future agricultural profits is sensitive to the four pre-determined groupings currently used in the literature. The results indicate that accounting for grouping uncertainty greatly increases the confidence interval around projected climate impacts. In addition, we do not find that one type of grouping is superior to any other. We suggest two potential solutions and emphasize the importance of explicitly controlling for grouping uncertainty in future studies.