Drift algae may represent an overlooked subsidy that could alter standing stocks of algae and herbivores at downstream sites, with important consequences for coral reef systems.  At Moku o L‘oe in Kāne‘ohe Bay, Hawaii low inputs (<22 ± 14 g trap^-1 day^-1) of drift algae precluded algal accumulation on the reef slope, but accumulation at sites with substantial drift input (>621 ± 208 g trap^-1 day^-1) was modified by other factors.  Caged algae showed no in situ growth rate differences among sites, but six out of seven tests suggested that herbivores were modifying algal accumulation.  Both 1) total and 2) preferred drift algal inputs decreased towards the reef crest at a low-accumulation site.  Both types of algae had higher bottom cover at a high- vs. a low-accumulation site, but 3) although cover of preferred algae increased towards the inner reef flat at both high- and low-accumulation sites, 4) cover of non-preferred algae increased towards the reef crest at the high-accumulation site.  5) Herbivory decreased away from the reef crest, and was greater at a low- vs. a high-accumulation site.  Finally, 6) algal cover and herbivore biomass were negatively correlated at the northern end of the island, but 7) not at the island-wide scale.  Herbivore biomass was positively correlated with drift input, but not with coral or algal cover.  These results suggest algal accumulation at the reef crest is a function of drift algal input and herbivore pressure, and that drift input may be an important trophic subsidy for herbivorous fishes.  In general, drift algal subsidies to downstream areas may have important and underappreciated ecological implications for the coral reef system.