Canopy-forming algae occur across of range of energy environments (i.e., wave sheltered to exposed coasts) where disturbances are frequent (i.e., gap formation) and benthic patterns largely reflect variation in post-disturbance processes. Disturbances vary in extent (area affected) and intensity (degree of damage), and this may affect recolonisation at local scales. On an open oceanic coast, we tested whether habitat structure (patches of canopy algae) differed between heavy and relatively lighter wave exposure (sheltered vs. exposed sides of islands), and whether wave exposure affected the response of prominent habitat-formers to varying disturbance regimes (different sizes of partial and complete canopy removal). Observations of naturally occurring patterns showed sheltered coasts to be characterised by small patches of fucoids, whereas exposed coasts were characterised by large patches of kelp. Canopy-gaps were larger at exposed than sheltered coasts, and mixed canopies constituted > 24% of the subtidal rocky habitat independently of wave exposure. Experimental disturbances showed the local density of kelps to affect recovery through greater recruitment to partial clearings (80% canopy removal). Fucalean algae, on the other hand, mainly recruited into complete clearings (100% removal), but when their recruits were abundant, they also recruited into partial clearings. The covers of filamentous, turf-forming algae increased in all clearings, and more so at exposed than sheltered sites. Extent of disturbance had no detectable effect on recolonisation by canopy-forming algae across the scales examined (i.e., 1.5 m, 3 m diameter loss of canopy). Recolonisation varied among islands kilometres apart, and correlations (r > 0.85) between cover of canopies and cover of their recruits in clearings at the scale of sites, suggested that differences in propagule supply could account for variation in patterns of recolonisation at scales of kilometres. There was no evidence to suggest that the effect of disturbance depended on wave exposure within the range of exposures tested in this study (i.e. open coasts). We recognise that wave exposure can be fundamental to habitat structure of subtidal rocky coasts, but we suggest that its influence may be mediated by the biological setting (e.g., canopy composition). ; http://www.elsevier.com/wps/find/journaldescription.cws_home/600318/description#description ; Thomas Wernberg and Sean D. Connell