Abstract: We address whether vascular plants can serve as a surrogate for ants in fine-scale assemblage studies by investigating assemblages of four plots of a dry steppe habitat in eastern Austria. We found that ants drew another picture than plants with respect to evenness, assemblage similarity, meta-similarity between single assemblage similarity matrices, and spatial species turnover. Ant data was more robust to data transformation than plant data. Differences between plant and ant assemblages levelled when data were reduced to presence-absence mode. We suspect that, in general, the correlation between plant and ant assemblages may be higher for coarse-scale studies involving pronounced ecological differences between habitats.