The species composition of macrofauna associated with floating seaweed rafts is highly variable and influenced by many factors like spatial and temporal variation, period since detachment and probably also the seaweed species. The presence of seaweed preferences was assessed by a combination of in situ seaweed samplings and multiple-choice aquarium experiments in a controlled environment, using the seaweed-associated grazing organisms Idotea baltica and Gammarus crinicornis . Results from the sampling data confirm that the seaweed composition influences macrofaunal species composition and abundance: samples dominated by Sargassum muticum displayed higher densities but lower diversities compared to samples dominated by Ascophyllum nodosum and Fucus vesiculosus . Seaweed preference was also apparent from the multiple-choice experiments, but did not exactly match the results of the community analysis: (1) I. baltica had high densities in seaweed samples (SWS) dominated by F. vesiculosus and A. nodosum , while in the experiments, this isopod was most frequently associated with Enteromorpha sp. and F. vesiculosus , and fed mostly on S. muticum , A. nodosum and Enteromorpha sp.; (2) G. crinicornis had high densities in SWS dominated by F. vesiculosus , while in the experiments, this amphipod was most frequently associated with S. muticum , but fed most on A. nodosum and F. vesiculosus . It is clear from the laboratory experiments that preference for habitat (shelter) and food can differ among seaweed species. However, food and habitat preferences are hard to assess because grazer preference may change if choices are increased or decreased, if different sizes of grazers are used, or if predators or other grazers are added to the experiments. The effects of seaweed composition may also be blurred due to the obligate opportunistic nature of a lot of the associated macrofaunal species.