Despite their central roles in ecology, patterns and scales of variation in the distribution of species and their densities are seldom satisfactorily described for most ecosystems. This is mainly due to spatial structures being scale-dependent across broad ranges of scales with partly different processes operating at different scales. In this study, scale-dependent distributional patterns in soft-sediment macrofaunal communities of naturally low species richness were studied on the Aland Islands, northern Baltic Sea, in 3 model areas (I, II and III) representing different environmental regimes with regard to openness (water renewal rate), depth, and organic and nutrient content as well as an increasing intrasite variability. A systematic grid design comprising 64 biotic and 16 abiotic sampling points, hierarchically arranged at 3 (at distances 10, 100 and 1000 m) and 2 spatial scales (100 and 1000 m), respectively, was used in each area. The intrasite distribution of macrofauna was predictably the most uniform in Area I, the least variable environmentally, followed by Area II, which was intermediate with respect to both abiotic and biotic variability, and then Area III, which was always the most variable area. This was true when analysed by both univariate and multivariate statistical means. When examining the contribution of each spatial scale to the total variance, most variability was observed at the smallest sampled level (10 m) for the majority of the studied variables. Despite a high overlap in species composition, the areas differed significantly in animal community structure. This was driven by both (1) the top-dominant species (Macoma, Mytilus, Monoporeia, Oligochaeta, Marenzelleria) and (2) the proportion and distribution of rare species. These systems with low diversity showed clear species distribution patterns driven by local environmental conditions, but also showed a substantial biotic component seen as the proportion and role of rare species and the high spatial variability observed at the smallest scale. Our understanding of spatial variability (or 'stability') is thus influenced by both the habitat and the species composition. The results are valuable for the refinement of sampling design, improved interpretation of monitoring programs (including detection of anthropogenic impact) and incorporation of scaling issues into questions of marine biodiversity and conservation.