Bioassays using the nonforced exposure approach have been shown to be a relevant tool that might complement the traditional ecotoxicological risk assessment. Because the nonforced exposure approach is based on spatial displacement of organisms and the consequent habitat selection processes, the population density might play an important role in the decision to avoid or prefer an ecosystem. Therefore, the aim of the present study was to assess if the avoidance response to contamination, measured in a nonforced exposure system, is density‐dependent and how determinant contamination could be for the habitat‐selection process in comparison with the population density. The freshwater shrimp Atyaephyra desmarestii was exposed to a copper gradient in a nonforced exposure system formed by 7 interconnected compartments (total volume 600 mL), which contained different copper concentrations. The density treatments used were 3, 5, and 10 organisms per compartment, corresponding to 0.5, 0.8, and 1.7 organisms per 100 mL, respectively. Clearly, the avoidance response to copper was more intense in the population with the lower density: the highest population density showed the lowest avoidance. The concentrations that triggered an avoidance of 50% of the population were 47, 134, and 163 μg L−1. In summary, it was observed that shrimps were able to detect and avoid potentially toxic copper concentrations but that the avoidance response was affected by population density. Environ Toxicol Chem 2018;37:3095–3101. © 2018 SETAC