Understanding the dynamics of epibiosis is fundamental to understanding the role of biological interactions in the functioning of marine ecosystems. At many coastal sites, barnacles are abundant as epibionts on the shells of mussels. As they approach the shore, cypris larvae explore the environment, actively seeking suitable settlement sites and rejecting unsuitable substrata. Suitability involves micro-features of the surface, including its texture and contours, chemical signals and the presence of conspecifics. To investigate whether epibiosis on mussel shells indicates a preference for this substratum on the part of the barnacle Chthamalus dentatus, artificial plates were deployed at two rocky shore sites on the south coast of South Africa. These plates offered settling cyprids four habitat choices: a live mussel, the shell of a recently dead mussel, a fine resolution resin replica of a mussel shell to mimic architectural micro-surface, without the chemical characteristics of natural shells and a rock mimic, a surface covered in a film of hard plastic that resembled natural rock surface as closely as possible. Settling surfaces were photographed monthly, and new barnacles were counted and allocated to size classes, so that survival could be estimated over a period of several months. Barnacles showed a clear preference for the rock mimic surfaces, which on average supported double the number of settlers found on mussel replicas. These in turn were significantly higher than the numbers on live and dead mussel shell surfaces, implying that there are features of both dead and live mussel shells that deter barnacle settlement. In contrast, final abundances of adults showed no significant effect of settling surface. The results suggest that during settlement, cyprids avoid chemical cues from mussel shells that persist even after the death of the mussel. This suggests that the high numbers of barnacles often found on the surface of live mussels could be due to saturation of substratum.