Macroinvertebrate community responses to copper (Cu) in laboratory streams receiving dechlorinated tap water and field streams receiving natural river water were compared. Since both tap water and natural river water were taken from the same source in these experiments, several of the important water quality parameters known to influence heavy metal toxicity were similar. Despite the fact that field streams were initially dominated by metal-sensitive Ephemeroptera, effects of Cu were greater in the laboratory. After 10 days of exposure to Cu (11.3 (μg/L), macroinvertebrate abundance was reduced by 75% in laboratory streams compared to 44% in field streams. In the field, the number of taxa was reduced by 10% in treated streams compared to 56% in the laboratory. The response of dominant taxa to Cu exposure was also greater in the laboratory. In the field, abundance of metaltolerant caddisflies (Hydropsychidae:Trichoptera) was similar in control and treated streams (11.3 μg/L). These organisms were reduced by 71% in laboratory streams at similar Cu levels. The greater impact of Cu observed in laboratory streams may have resulted from the inability of certain taxa to acclimate to laboratory conditions. The usefulness of community-level toxicity tests in the laboratory may be limited due to the overestimation of metal toxicity. We recommend that future experiments be conducted in field mesocosms receiving water directly from the system under investigation. © 1990 Springer-Verlag New York Inc.