The goal of parasite epidemiologists is to understand the factors that determine host infection levels. Potential infection determinants exist at many scales, including spatial and temporal environmental variation, among-host differences, and interactions between symbionts infecting the same host. All of these factors can impact levels of parasitism, but frequently only a subset is considered in any host-parasite system. We examined several potential determinants of pinworm infection in wild Australian cockroaches (Periplaneta australasiae) from multiple biological scales: (1) habitat; (2) season; (3) cockroach body size, developmental stage, and sex; and (4) interactions between 2 pinworm species (Leidynema appendiculata and Thelastoma sp.). Over 1 yr, we collected 239 cockroaches from 2 separate rooms in an Illinois greenhouse. We used generalized linear mixed-effects models (GLMMs) to evaluate simultaneously the influence of these factors on pinworm abundance, and nearly all had significant effects. Overall, the abundance of L. appendiculata was greater than Thelastoma sp., but the relative abundance of the 2 species was reversed in each room (i.e., a taxon × habitat effect). Abundance varied over 4 trapping seasons and increased with cockroach size. Adult cockroaches had more pinworms than nymphs, and there was also a significant taxon × stage effect: adult cockroaches had fewer pinworms than expected for their larger size, and this reduction was greater in Thelastoma sp. than in L. appendiculata. Cockroach sex had no effect on infection. Although females had more worms than males, this difference could be explained by the larger size of females. Finally, after controlling for all other potential determinants of infection, we found a strong negative association between Thelastoma sp. and L. appendiculata; cockroaches tended to be infected with either 1 pinworm species or the other. Our work underscores the importance of measuring potential determinants of infection from as many scales as possible. Such approaches are necessary to unravel the complexities of host-parasite interactions.