We investigated the effects of differing spatial scales of seagrass habitat architecture on the composition and abundance of settling bivalves in a sub-tropical seagrass community. The density of newly settled bivalves was generally greater at Thalassia testudinum grass bed edge (<1 m) compared to interior portions of the bed (>10 m). Deviation from this generalized pattern occurred when high densities of newly settled tulip mussels (Modiolus americanus) were recorded from the interior of the meadow, associated with aggregations of adult mussels. Bivalve settling densities appear to reflect settlement shadows of passively delivered larvae, bedload transport of newly settled individuals from unvegetated regions, as well as gregarious settlement among adult conspecifics. We also investigated the impact of seagrass patch shape and size on settlement by using artificial seagrass units (ASU) in separate short-term and long-term experiments. We found a positive relationship between ASU perimeter and bivalve abundance, suggesting that larval encounter rates with seagrass habitat may determine initial settlement patterns. Using ASUs we also investigated the relative role seagrass epiphytes play in determining the density of settling bivalves. Results showed greater settling densities where epiphytic secondary structure was elevated compared to controls, and bivalve density was significantly greater when ASUs were fouled with a natural community of epiphytes, suggesting that both microstructure and biofilms positively influenced bivalve settlement. We conclude that structural components of seagrass habitats increase bivalve settlement at multiple spatial scales, including epiphytic micro-structure, small-scale patch shape and size, and large-scale within habitat differences.