Numerous seaweeds produce secondary metabolites that are allelopathic to corals. To date, most of the compounds identified in this interaction are lipid-soluble instead of water-soluble. Thus, understanding whether these compounds are stored internally where they would not contact corals, or occur on external surfaces where they could be transferred to corals, is critical to understanding seaweed-coral interactions and to informing realistic experiments on chemically mediated interactions. We conducted field experiments assessing the effects of lipid-soluble extracts from macroalgal surfaces alone versus total lipid-soluble extracts from both internal and external tissues on the coral Pocillopora verrucosa. Extracts of the red algae Amansia rhodantha and Asparagopsis taxiformis, the green alga Chlorodesmis fastigiata, and the brown alga Dictyota bartayresiana suppressed coral photochemical efficiency; in these bioactive species, the total lipid-soluble extracts were not more potent than surface-only extracts despite the concentration of total extracts being many times greater than surface-only extracts. This suggests that previous assays with total extracts may be ecologically meaningful, but also that future assays should be conducted with the simpler, less concentrated, and more ecologically relevant surface extracts. Allelopathic effects of As. taxiformis and C. fastigiata were significantly greater than the effect of D. bartayresiana, with effects of Am. rhodantha intermediate between these groups. Neither surface-only nor total lipid-soluble extracts of the seaweed Turbinaria ornata were allelopathic, and its lack of potency differed significantly from all other species. Our results suggest that lipid-soluble, allelopathic compounds are usually deployed on seaweed surfaces where they can be effective in surface-mediated interactions against other species.