Almost all terrestrial and aquatic assays that are accepted as standardized, or that have been proposed, involve the exposure of confined organisms to toxicants. If avoidance (sensu evasion, displacement) of contaminants occurs in real situations, then bioassays involving forced exposure severely underestimate pernicious effects of contamination. Two main objectives were achieved in this study: To verify the occurrence of avoidance of copper by cladocerans; to determine the association between avoidance and other toxicity endpoints (lethality and feeding depression), and therefore, to ascertain if fully acclimated individuals from a reference field population exhibited a genetically determined avoidance markedly different from those belonging to a historically metal-stressed population. Twelve cloned lineages of the cladoceran Daphnia longispina, collected from two field populations, were selected according to their lethal sensitivity to copper and acclimated to controlled conditions for more than 30 generations. A 1.1-m test chamber with five compartments was built, allowing the establishment of a dissolved toxicant gradient and the free movement of individuals. In the absence of any toxicant, juveniles from each cloned lineage distributed themselves randomly along the test chamber and furthermore, no significant differences were observed between the two replicates, attesting the repeatability of this novel assay. All lineages showed significant avoidance to copper when exposed to a gradient from 3 to 87 μg/L. The most sensitive lineages to lethal levels of copper began avoiding this metal earlier than resistant ones. An intense association was observed between other endpoints and avoidance; furthermore, avoidance was determined to be much more sensitive than lethality. Therefore, avoidance assays should be recommended as a complementary tool in ecological risk assessments and effluent biomonitoring because such assays can provide cost-effective and ecologically relevant information.