Much literature is dedicated to the study of density dependence in host-parasitoid systems. The theoretical literature identifies two potential stabilization mechanisms relating the response of the parasitoid to host density. One is a behavioral response that occurs at a local level, and the other is a demographic response that occurs at a larger spatial scale with heterogeneous patches of host concentrations. In a series of field trials at the small spatial scale (tens of meters) and a 10-mo census at the large spatial scale (hundreds of meters), we examined the attack rates of Pseudacteon sp. phorid flies on Azteca instabilis F. Smith ant hosts. At the local scale, we examined phorid attack rates on ants placed at increasing distances from ant nests (0-12 m) and on different densities of ants (1-50 workers) placed near the nests. At the large scale, we placed whole body extracts of A. instabilis in areas with several A. instabilis nests and in areas >100 m from the nearest nest. For all observations, we recorded the time of the first phorid attack and the number of phorid attacks in a defined time period. We found clear density-dependent responses at both scales. Phorid attack rates were highest within 2 m of A. instabilis nests and at ant densities >25. At the larger scale, phorid attack rate was greater in areas with A. instabilis nests, although this result was obscured during the dry season when the population of parasitoids is dramatically reduced. We propose several behavioral and population-level responses that may contribute to the observed results and propose several testable hypotheses. We conclude that, in this system, density dependence does happen through a behavioral mechanism of the phorid actively seeking concentrations of ants, but also, a population level response is likely caused by the significant difference in phorid attack rates in areas far from ant nests compared with areas nearby to nests.