ABSTRACT. Gastropod loconlotion typically lnvolves the deposition of a mucus trail. This trail can be energetically costly, and its producer could defray this cost if the trail were to perform some post-deposition function. Here we report laboratory experiments aimed at assessing the potential role in nutrition of the mucus trail of the common intertidal periwinkle Littorina littorea (L.). Mucus trails bound more microalgal cells from suspension than d~d a glass substratun~, microalgal densities increasing with period of exposure to the suspension up to -5-8 h. After this time adhesion was less (but still greater than to glass). Amphora coffeaeformis (pennate diatom) adhered in greater densities than did Tetraselmis suecica (flagellate prasinophyte). Mucus trails containing microalgae of both species (-50 to 100 cells mm-'; approximatly the conditions on-shore) induced a greater degree of trail-following than did bare trails. Individuals followed conspecific trails longer than they did their own trails. Winkles moved significantly slower on bare mucus trails (mean = 0.35 mm S-') than on glass (0.68 nlnl S-'), though the addition of nucroalgae to mucus increased the speed of the winkles, A. coffeaeformis significantly so. Feeding rate (rate of radular rasping) was also significantly increased on trails containing A. coffeaeformis (mean = 17.8 bites min-l) and T suecica (12.9 bites min-') in comparison to control trails (4.3 bites min-'), where a form of searching behaviour occurred. Microalgae embedded in mucus were seen to enter the mouth. Only 3 out of 40 periwinkles showed any radular activity on glass. Following the passage of a winkle, the density of A. coffeaeformis was reduced by -38% and that of T. suecica by 43 %. Winkles can clearly exploit food (rnicroalgae) in conspecific mucus trails and in doing so modify their trail-following and feeding behaviour. Thus trail-following seems inextncably linked to nutrition. Since much of the intertidal is likely to be covered with a layer of mucus-or its degradation products-those experiments on trail-following that used 'clean' substrata were not representative of the field. Distribution patterns of both L. littorea and benthic microalgae might be shaped by the ability of mucus tralls to bind rnicroalgae and by their subsequent exploitation by the grazer.