Abstract Successful fisheries management is underpinned by an understanding of the processes that underlie the population dynamics of exploited stocks. This study considered the effects of experimental harvesting on recruitment of Mytilus galloprovincialis along the west coast of South Africa, where harvesting of this alien species is being contemplated. In particular, the role of settlement habitat availability in the form of adult mussels was analysed. To track the effects of a spectrum of harvesting intensities, five treatments were implemented: F = 0 (i.e. a control), F = 0.3, F = 0.6, F = 0.9 and F = 1. At these harvesting intensities 0%, 30%, 60%, 90% or 100% of mussel biomass was removed respectively at the outset of the experiment. A significant negative exponential relationship ( p < 0.01) was found between M. galloprovincialis recruit density and harvesting intensity, with intensities greater than F = 0.3 dramatically reducing recruitment. This pattern was recorded throughout the intertidal zone and remained temporally constant over 2 years. Significant positive linear relationships ( p < 0.01) between recruit density and adult mussel biomass or density indicate a strong correlation between availability of settlement habitat and recruitment. It is likely that the high recruit density recorded at low harvesting intensities (2000–20 000 per 0.01 m 2) exceeds the level required for population maintenance. However, if settlement habitat is eliminated or significantly reduced, as was achieved by F = 0.3 or above, recruitment may become limiting. Thus, to protect stock replenishment, harvesting of M. galloprovincialis in this region should take place at intensities less than F = 0.3.