Ocean surgeonsAcanthurus bahianusand doctorfishA. chirurgusdisplayed no preference between food items containing toxic dinoflagellates associated with ciguatera poisoning and nontoxic food when given a choice for up to 42 consecutive days. Two clones of laboratorycultured dinoflagellatesGambierdiscus toxicusand one clone ofProrocentrum concavumwere used in separate trials.Gambierdiscus toxicusclone SIU-350 was toxic to both fish species after 5–8 d at a daily ration of 2.5 mg of cells per food item. Clinical signs were similar to those previously reported for blueheadsThalassoma bifasciatumfed the same clone, Clinical signs became less pronounced with time, but recurred with equal intensity when the ration was doubled. No difference in food consumption between toxic and nontoxic food types was observed. Fish fedG. toxicusSIU-175, previously shown to be considerably more toxic to miceMussp. and brine shrimpArtemiasp. than toG. toxicusSIU-350, remained asymptomatic at daily rations as high as 7.5 mg of cells per food item. No signs of toxicity were observed in adult brine shrimp that consumed ground muscle and liver tissue from fish fed eitherG. toxicusclone, indicating that ciguatoxicity had not been induced. Results suggest that herbivorous fish may acclimate to, rather than avoid, macroalgae harboring epiphytic toxic dinoflagellates. Attempts to directly relate these findings to nature should be made cautiously because differences in toxic properties between cultured and wild dinoflagellate cells cannot be discounted.