Understanding the ecological function of species and the structure of communities is crucial in the study of ecological interactions among species. For this purpose, not only the occurrence of particular species but also their abundance in ecological communities is required. However, abundance quantification of species through morphological characters is often difficult or time/money consuming when dealing with elusive or small taxa. Here we tested the use of next-generation sequencing (NGS) for abundance estimation of two species of feather mites (Proctophyllodes stylifer and Pteronyssoides parinus) under five proportions (16:1, 16:4, 16:16, 16:64, and 16:256 mites) against a mock community composed by Proctophyllodes clavatus and Proctophyllodes sylviae. In all mixtures, we retrieved sequence reads from all species. We found a strong linear relationship between 454 reads and the real proportion of individuals in the mixture for both focal species. The slope for Pr. stylifer was close to one (0.904), and the intercept close to zero (-0.007), thus showing an almost perfect correspondence between real and estimated proportions. The slope for Pt. parinus was 0.351 and the intercept 0.307, showing that while the estimated proportion increased linearly relative to real proportions of individuals in the samples, proportions were overestimated at low real proportions and underestimated at larger ones. Additionally, pyrosequencing replicates from each DNA extraction were highly repeatable (R = 0.920 and 0.972, respectively), showing that the quantification method is highly consistent given a DNA extract. Our study suggests that NGS is a promising tool for abundance estimation of feather mites' communities in birds.