The strengths of biological processes (e.g., larval settlement, competition, predation, distribution of food resources) and physical factors (e.g., desiccation, freezing, salinity fluctuations) can be correlated with tidal height and may contribute to the vertical zonation of many rocky intertidal organisms. Habitat structural complexity has also been shown to influence the density and diversity of marine organisms and could contribute to vertical zonation where the level of complexity varies significantly with tidal height. To test the hypothesis that vertical zonation of brachyuran crabs was related to structural complexity, shelter was experimentally manipulated at 2 heights at a rocky intertidal site in Tanabe Bay, Japan. The densities of four brachyuran species, varying in size and life history type, were compared before and after the manipulation. Removal of shelter negatively affected all species, regardless of tidal height. Doubling shelter positively affected Hemigrapsus sanguineus de Haan and Leptodius exaratus Milne Edwards, especially at the tidal heights where they were initially more abundant. Two smaller more opportunistic species, Gaetice depressus de Haan and Acmaeopleura parvula Stimpson, did not react positively to shelter increases and each showed unique responses to the experimental manipulations. Differences in the amount of structural complexity in the upper and lower intertidal may be responsible for conflicting literature reports of the vertical distribution of Hemigrapsus sanguineus. Information gathered about the habitat requirements of this invasive grapsid crab will be useful for monitoring its impact and spread along the eastern coast of the United States. (C) 2000 Elsevier Science B.V. All rights reserved.