Ngative interaction between individuals is a distinctive feature of populations. The question is addressed whether the population response to toxicants (e.g., the pyrethroid pesticide fenvalerate) is altered by the strength of negative intraspecific interaction. Larvae of the trichopteran Limnephilus lunatus Curtis were exposed for 1 h to fenvalerate, then larval survival and adult emergence were monitored in outdoor stream microcosms over eight months. Degree of interaction was varied by changing larval densities. No survival was observed after 15 d at 100 μg/L. Survival decreased slightly at 10 μg/L when strong interaction was present. Survival in the range of 0.001 to 1 μg/L was similar to that in the control, regardless of the strength of interaction. In contrast, chronic effects differed widely according to the degree of interaction. Weak interaction resulted in high survival of controls, a lowest-observed-effect concentration (LOEC) (mortality) of 0.1 μg/L, a delay in development at 0.001 to 0.01 μg/L, and reduced weight of adults from 0.01 μg/L on. Strong interaction resulted in low survival of controls, a LOEC (mortality) of 100 μg/L, and a delay in development at 0.1 μg/L. The suggestion was made that these differences in chronic effects are caused by a toxicant-induced reduction of negative intraspecific interaction at high concentrations (≥1 μg/L). Hence, direct effects of the toxicant are partly compensated because of indirect reduction of negative interaction compared to the control. Interaction is not decreased at low concentrations (≤0.1 μg/L) and, thus, effects of the toxicant are not compensated. Accordingly, the response to toxicants at the population level when negative interaction is present differs from the response at individual level.