[Truncated abstract] The ecological effects of marine protected areas (MPAs) in temperate ecosystems are poorly understood relative to their tropical counterparts. The limited number of rigorous empirical studies supporting existing theoretical models, increasing public awareness of the importance of marine conservation strategies and legislative requirements to review management effectiveness provide further impetus to study temperate MPAs. Investigations should consider confounding effects of natural variability if MPA effects are to be clearly demonstrated. This research helps to address these needs by investigating the short term effects of sanctuary zones (no-take MPAs where fishing is prohibited) and wave exposure at Marmion Marine Park, Western Australia. The three sanctuary zones at Marmion Marine Park are extremely small (0.061 0.279 km2) compared to most reported in the literature. The sanctuary zones are nested within a larger, fished zone (94.95 km2). The sanctuary zones have been protected from fishing since the year 2000. A post-hoc, asymmetrical sampling design was used in this study and involved surveys of fishes, mobile invertebrates and macroalgae at one sanctuary zone and two fished sites (controls) at each of three successive, subtidal reef lines. The three reef lines are exposed to a gradient in wave energy. The size structure and abundance of the heavily exploited Panulirus cygnus (Western Rock Lobster) were positively affected by protection from fishing in sanctuary zones, despite the highly mobile nature of this migratory species. The mean abundance of legal size lobsters was higher in sanctuary zones compared to fished sites during an interannual study (2003, 2005 and 2006). The total abundance of lobsters and the mean abundance of legal size lobsters were higher at inshore and offshore sanctuary zones compared to fished control sites during a 2005/2006 fishing season study. These zoning effects did not vary with the time of survey. ... Furthermore, the abundance of large lobsters in sanctuary zones decreased with the duration of the 2005/2006 fishing season. Similarly, it is likely that sanctuary zones are too small relative to the movement of fishes to adequately protect stocks of some targeted species. The small sanctuary zones at Marmion are unlikely to offer protection to highly mobile species over the long term. And finally, ecological assemblages within each level of wave exposure are distinct. Consequently for each assemblage type, the current reserve design does not include replication of sanctuary zones and does not offer any 'insurance' in the event of isolated impacts affecting a particular zone. This study has identified the benefits and deficiencies of the design and function of small no-take temperate MPAs in Western Australia. An increase in the size and number of sanctuary zones within each wave exposure level will help to address the v shortfalls of the zoning scheme and enhance the conservation benefits of management at Marmion Marine Park. More generally, this study demonstrates that the mobility of the species to be protected from fishing should be considered when designing MPAs. Lessons learned from this work will be beneficial for the future management and conservation of resources in the region and elsewhere.