In size—structured populations interspecific competitors often have been suggested to have their greatest impact on the juvenile life stage. We hypothesized that roach (Rutilus rutilus), by virtue of their superior foraging ability on zooplankton, would cause juvenile perch (Perca fluviatilis) to shift from feeding on zooplankton to macroinvertebrate prey, the prey consumed by adult perch. This should potentially increase size—class competition in perch, producing a competitive bottleneck at the macroinvertebrate feeding stage. To test this, we experimentally examined how roach density affected both the diet and growth of young—of—the—year (0+) and 1—yr—old (1+) perch as well as how it affected the zooplankton and benthic macroinvertebrate resource levels in 10 enclosures in a pond over a 3—mo period. The densities of 0+ and 1+ perch were held constant between enclosures, where the density of roach was varied over five density levels. A significant effect of roach density on zooplankton (cladoceran) biomass was observed over the experimental period. By the end of the experiment, the biomass of zooplankton was low in all enclosures except those without roach, and the mean size of two zooplankters, Bosmina and cyclopoid copepods, decreased with increasing roach density. In contrast, the total biomass of macroinvertebrates in the enclosures did not change with increasing roach density. There was, however, evidence that chironomids were released from predation by Sialis due to an increased consumption of Sialis by 1+ perch. As predicted from our hypothesis, 0+ perch shifted from pelagic zooplankton to benthic macroinvertebrates as roach density increased. The roach density at which 0+ perch shifted to feed on benthic prey decreased over the experimental period, but occurred at the same density of pelagic prey. The proportion of zooplankton in the diet of 1+ perch and roach also decreased with increasing roach density, with 1+ perch totally dropping zooplankton from their diet and roach shifting to plant to detrital material. Over time we observed that roach increased its utilization of plant and detrital material. We found density dependence in growth rates of roach and 0+ perch as well as 1+ perch. The growth data suggested that the main competitive effect of roach occurred during the first half of the experiment. This was related to roach's high consumption of plant material at all roach densities at the end of the experiment and to the low water temperature during the second half of the experiment.