Four artificial reef designs and a natural control reef were compared to determine their relative value as a fisheries enhancement tool in tropical waters. Reefs were comparably sized and designs ranged from haphazardly dumped scrap materials to carefully deployed, specifically designed concrete modules. All reefs were situated in the same biotope and depth, were similar sized and were located within 1 km of each other. Fish density, community structure and standing crop estimates on each reef type were made visually. Reefs composed of haphazardly dumped scrap materials (automobile shells and surplus concrete pipe) provided the poorest enhancement. These reefs were also highly unstable and exhibited low life expectancies. Reefs composed of modules of scrap automobile tires set in concrete bases and dumped haphazardly showed moderate enhancement that varied with the degree of module dispersion. Due to their high mass to volume ratio, modules of this design were relatively stable but the design precluded effective stacking, resulting in low relief structures. Significantly greater enhancement effects(e.g., mean standing crop, mean size per fish and mean number of species) were attained on a small reef constructed of 42 open framework concrete cube modules arranged to provide maximum refuge space for fishes. The open framework concrete cube module was engineered for a long life expectancy and stability in high energy environments. The data suggest that haphazard deployment of materials provided significantly poorer enhancement relative to a reef constructed of designed modules assembled into a specific configuration.