The bloodworm, Arenicola loveni, is commonly used as bait by fishers and may be vulnerable to local extinction owing to its K-selected life history strategies and overexploitation. Accurate population data that can inform management is, however, outdated, whereas demand for bloodworm continues or is even increasing. This study provides genetic data on the population structure of A. loveni sampled at seven sites in the Western Cape and one in the Eastern Cape Provinces of South Africa. Mitochondrial cytochrome oxidase I (COI) data indicate that the population forms two lineages that should be managed separately: a western lineage that contains mostly samples from Saldanha Bay and a southern lineage that contains samples from the south coast (Muizenberg to Swartkops). High haplotypic, but low nucleotide diversity for the nuclear internal transcribed spacer (ITS) and COI genes, suggest that populations from both lineages have a recent common ancestry. Low pairwise Fixation index (F_st) values among most populations on the south coast, and common haplotypes shared among localities, indicate some degree of connectivity among populations. The presence of many private haplotypes at each site, however, indicates that local populations also rely heavily on regional recruitment. Gene flow among populations, and the absence of this between the western and southern lineages, is most likely attributed to larval dispersal facilitated by the predominant oceanic circulation patterns. These results suggest that although individual populations may be protected against local extinction through recruitment from elsewhere, reliance on local recruits may render populations vulnerable should baiting pressure increase.