Red-plumaged birds (red species) may behave more visibly in the presence of a red-orange safety vest that field observers must often wear, but species lacking red plumage (non-red species) may behave less visibly (i.e., the species-confidence hypothesis). Such responses may bias many common counting techniques. We experimentally tested if observer clothing color could affect richness and abundance variables derived from point counts. During summer and winter, we surveyed point-count stations while alternately wearing a red/orange vest (vest counts) and not wearing a vest (control counts). Winter estimates of species richness, total abundance, abundance of non-red species as a group, and the White-breasted Nuthatch (Sitta carolinensis) were significantly lower when the vest was worn, but red species were not significantly different. We did not observe any significant effects of the vest during summer counts. Power was high for community-level variables and intermediate for non-red species, but low power and marginally significant results suggested that we might have failed to detect effects on red species. Several factors may account for why we observed effects on non-red species during winter but not summer: species composition varied between seasons, dense summer foliage may have obscured the vest from view, or territoriality and nesting activities may have dampened avoidance behaviors during summer. Although effects of clothing color were significant for the winter season, managers and biologists should avoid risk of bias by restricting observers to dark or drab colored clothing for all seasons and for all avian sampling techniques, not just winter point counts.