The potential for enhancing fish abundance, species richness, and biomass on artificial reefs was examined by attaching floating attractants and manipulating structural complexity of small concrete reefs each approximately 1.3 m in diameter, 1 m high. Experimental design consisted of a comparison of fish assemblages among three treatments (10 replicate, hemisphere-shaped reefs each): 10-m floating line attached (Streamer); concrete block in the central void space (Block); and no floating line or concrete block (Control). Reefs were deployed on sandy substrate at 20-m depth off Fort Lauderdale, Florida, USA. Divers recorded fish census data on slates 18 times over 24 months. Species composition, numbers of individuals per species, and estimated total length (TL; by size class: <5, 5–10, 10–20, and >20 cm) for all fishes within 1 m of each reef were recorded. Size classes were used to calculate fish biomass. There was a significant difference among treatments. Block reefs had higher numbers of individuals, species, and biomass than Streamer or Control reefs (p<0.05). With one exception, Streamer reefs did not differ from Controls for any of the parameters investigated (p>0.05). These results highlight the importance of structural complexity in artificial reefs designed to enhance fish recruitment, aggregation, and diversity. Copyright 2002 Published by Elsevier Science Ltd on behalf of the International Council for the Exploration of the Sea .