Measures of functional diversity are expected to predict community responses to land use and environmental change because, in contrast to taxonomic diversity, it is based on species traits rather than their identity. Here, we investigated the impact of landscape homogenisation on plants, butterflies and birds in terms of the proportion of arable field cover in southern Finland at local (0.25 km2) and regional (> 10 000 km2) scales using four functional diversity indices: functional richness, functional evenness, functional divergence and functional dispersion. No uniform response in functional diversity across taxa or scales was found. However, in all cases where we found a relationship between increasing arable field cover and any index of functional diversity, this relationship was negative. Butterfly functional richness decreased with increasing arable field cover, as did butterfly and bird functional evenness. For butterfly functional evenness, this was only evident in the most homogeneous regions. Butterfly and bird functional dispersion decreased in homogeneous regions regardless of the proportion of arable field cover locally. No effect of landscape heterogeneity on plant functional diversity was found at any spatial scale, but plant species richness decreased locally with increasing arable field cover. Overall, species richness responded more consistently to landscape homogenisation than did the functional diversity indices, with both positive and negative effects across species groups. Functional diversity indices are in theory valuable instruments for assessing effects of land use scenarios on ecosystem functioning. However, the applicability of empirical data requires deeper understanding of which traits reliably capture species' vulnerability to environmental factors and of the ecological interpretation of the functional diversity indices. Our study provides novel insights into how the functional diversity of communities changes in response to agriculturally derived landscape homogenisation; however, the low explanatory power of the functional diversity indices hampers the ability to reliably anticipate impacts on ecosystem functioning.