We investigated the non-consumptive effects (NCEs) of predatory dogwhelks (Nucella lapillus) on intertidal barnacle (Semibalanus balanoides) recruitment through field experiments on the Gulf of St. Lawrence coast and the Atlantic coast of Nova Scotia, Canada. We studied the recruitment seasons (May–June) of 2011 and 2013. In 2011, the Gulf coast had five times more nearshore phytoplankton (food for barnacle larvae and recruits) during the recruitment season and yielded a 58 % higher barnacle recruit density than the Atlantic coast at the end of the recruitment season. In 2013, phytoplankton levels and barnacle recruit density were similar on both coasts and also lower than for the Gulf coast in 2011. Using the comparative–experimental method, the manipulation of dogwhelk presence (without allowing physical contact with prey) revealed that dogwhelk cues limited barnacle recruitment under moderate recruit densities (Atlantic 2011/2013 and Gulf 2013) but had no effect under a high recruit density (Gulf 2011). Barnacle recruits attract settling larvae through chemical cues. Thus, the highest recruit density appears to have neutralized dogwhelk effects. This study suggests that the predation risk perceived by settling larvae may decrease with increasing recruit density and that prey food supply may indirectly influence predator NCEs on prey recruitment.