We studied spatial variation in the prevalence of the rhizocephalan barnacle Lernaeodiscus porcellanae on its porcelain crab host, Petrolisthes cabrilloi, at four southern California intertidal sites separated by only a few km. The prevalence of rhizocephalan externae varied significantly among sites in 2008-2009, with the southernmost site, White Point, consistently showing higher prevalence than the others. Externa prevalence was a good proxy of estimated true prevalence, i.e., the prevalence of rhizocephalans as a whole, not just those that had formed externae. We examined several hypotheses that might explain the observed spatial variation in prevalence. Host susceptibility to infection (indicated by the proxy of damage to host limbs, some of which are used to remove parasite infective stages), did not differ among sites. At all sites, prevalence was slightly higher in female crabs than in males, and the sex ratio at White Point was slightly female-biased while that at the other sites was male-biased; thus, among-site differences in sex ratio did contribute to observed variation in prevalence. However, most spatial variation in prevalence appeared to be due to the effect of host size. At all sites the probability of infection increased with increasing host size, and White Point crabs were on average much larger than crabs at other sites. Overall, the size-class distribution of host crabs explained 80.4% of the variation in prevalence of L. porcellanae. Larger P. cabrilloi have likely had greater opportunity to be infected by rhizocephalans, either because they are older, or because they have undergone more molts, during which they are especially vulnerable to infection. A deeper understanding of small-scale spatial variation in prevalence in L. porcellanae will require information on the causes of among-site variation in host population size structure.