The scleractinian coral Agaricia agaricites (Linnaeus) is a common component of reef systems throughout the Caribbean. The morphology of A. agaricites is extremely variable, including flat unifacial plates, upright bifacial plates, and encrusting forms. Transects conducted on the fore reef of Discovery Bay, Jamaica, indicated that the morphology of colonies growing on horizontal substrata was strongly related to depth. Colonies in shallower water (7–12 m) tended to encrust or form unifacial plates, while deeper colonies (20 m) were primarily upright and bifacial. Furthermore, 90% of all bifacial colonies were oriented directly perpendicular to (feeding surface facing) the dominant direction of flow. Bifacial colonies also tended to have larger corallites, and possessed ridges which angled upward away from the substratum. Measurements of flow conducted at this site indicated that ambient flow speeds generally decrease with increasing depth. A series of feeding trials was conducted in a laboratory flume over a range of flow speeds characteristic of those found on the reef (3–50 cm · s −1) to address the hypothesis that variations in colony morphology and orientation to flow represent mechanisms for maximizing particle capture. Upright bifacial colonies oriented perpendicular to flow fed at significantly higher rates than bifacial colonies oriented parallel to flow. Bifacial colonies were never, however, able to capture more particles per unit surface area than were unifacial plates, at any flow speed. In all colony morphologies tested, capture shifted from upstream to downstream areas with increasing flow speed, suggesting that feeding did not involve the inertial impaction of particles. Particle capture was highest at intermediate flow speeds, although horizontal-plating colonies were able to feed well over the entire range of flow speeds tested. Behavioral observations suggest that particle capture was aided by currents originating within the polyps, and apparently did not involve mucus entrapment, as previously suggested. Measurements of flow at points near the surface of colonies in the field indicated that flow conditions at 20 m roughly match flow conditions in the experimental flume, suggesting that the results of the feeding trials may be extrapolated to feeding in situ. The results thus suggest that colony morphology does not represent a mechanism for maximizing particle capture per unit tissue, but instead may have evolved as the result of other selective pressures such as spatial competition or gas exchange. Thus, by growing away from the substratum, colonies could potentially increase particle capture, and colony biomass, per unit of available substrate. The orientation of colonies into flow, however, does increase food capture, and may have arisen secondarily as a compromise between selective pressures.