Population density is a fundamental property of a species and yet it varies among populations of the same species. The variation comes from the interplay between intrinsic features of a species that tend to produce repeatable density values across all populations of the same species and extrinsic environmental factors that differ among localities and thus tend to produce spatial variation in density. Is inter-population variation in density too large for density to be considered a true species character? We addressed this question using data on abundance (number of parasites per individual host, i.e. equivalent to density) of fleas ectoparasitic on small mammals. The data included samples of 548 flea populations, representing 145 flea species and obtained from 48 different geographical regions. Abundances of the same flea species on the same host species, but in different regions, were more similar to each other than expected by chance, and varied significantly among flea species, with 46% of the variation among samples accounted by differences between flea species. Thus, estimates of abundance are repeatable within the same flea species. The same repeatability was also observed, but to a lesser extent, across flea genera, tribes and subfamilies. Independently of the identity of the flea species, abundance values recorded on the same host species, or in the same geographical region, also showed significant statistical repeatability, though not nearly as strong as that associated with abundance values from the same flea species. There were also no strong indications that regional differences in abiotic variables were an important determinant of variation in abundance of a given flea species on a given host species. Abundance thus appears to be a true species trait in fleas, although it varies somewhat within bounds set by species-specific life history traits.