In an experimental investigation of the relative importance of macrophyte shape and epiphyte cover in determining the distribution of epifaunal invertebrates within a Nova Scotian seagrass bed, colonization experiments were conducted, using plastic to simulate vegetation of shapes of differing complexity. 30 x 30-cm units were allowed to develop differing degrees of epiphyte cover, then deployed in the seagrass bed for 1 month. Both epiphyte cover and shape were important, but responses were highly species specific. Gammarus, Neopanope, Bittium, and Palaemonetes responded primarily to shape while Ampithoe, Chironomus, Leptochelia, and Mitrella responded more to epiphyte cover. In addition, laboratory experiments were conducted to investigate whether the efficacy of various macrophytes in providing refuge from predation could explain differential distribution of invertebrates. Predation by a small fish, Fundulus, and the crab Neopanope on a variety of invertebrates was not significantly affected by the presence or absence of artificial macrophytes. It was therefore concluded that while epiphyte cover is an important factor for some species, and shape for others, the influence of shape does not operate primarily through differential predation mortality.