A field experiment was conducted in spring 1991 to determine the utility of the cage exclusion method for experimentally evaluating imported natural enemies of the Russian wheat aphid, Diuraphis noxia (Mordvilko), and to make a preliminary assessment of the impact of resident natural enemies on this exotic pest. Screened cages were designed for total, partial, or no natural enemy exclusion. Uncaged plots of wheat served as a control. D. noxia densities increased significantly in the cages between the 1st and last sample dates (23 April and 14 May). Wheat in the uncaged plots appeared to be under stress that was unrelated to the level of D. noxia infestation. Aphid populations increased most rapidly in cages and reached the highest densities in the total-exclusion cages. Significantly fewer aphids occurred in the partial- and no-exclusion cages on the first and last sample dates than in the total-exclusion cages. Wheat damage increased as D. noxia density increased. Predators, mainly the coccinellids Hippodamia convergens (Guerin-Meneville) and Coccinella septempunctata L., increased throughout the experiment in the no-exclusion cages and in the uncaged plots. Mummies of aphidiid parasitoids were observed on only 1 date and these produced only secondary parasitoids. D. noxia densities varied little over time in the uncaged plots and may have been limited more by poor host plant quality than by natural enemies. Natural enemies appeared to have an impact on Russian wheat aphid populations, but the magnitude of this effect was difficult to determine because of possible cage effects. Therefore, we recommend that at least 1 other experimental evaluation technique be used in conjunction with the mechanical exclusion method.