Technical Abstract: The soybean aphid, Aphis glycines Matsumura, has become a serious pest of soybean in North America, and host-plant resistance is one potential management tool. In the current study, various F2-derived soybean selections with the Rag1 gene for resistance to soybean aphid were evaluated among F2-derived soybean selections without Rag1 and among contemporary soybean lines in a two-year field test. Overall, aphid levels per plant were over 10-fold greater in 2006 than in 2005, but lines generally performed similarly relative to one another between years with regard to aphid-infestation levels. In both years, the Rag1 selections ILL4, ILL27, ILL35, ILL37, ILL64RR, ILL76RR, and ILL77RR that contained Rag1 had the lowest mean number of soybean aphids per plant. However, in 2005, three selections putatively with Rag1--ILL26, ILL67RR and ILL87--had intermediate aphid infestation levels that were greater than those of other Rag1 selections, and in 2006 ILL26 and ILL67RR also had intermediate aphid levels that did not differ from all other lines. Irrespective of the Rag1 gene, all soybean lines tested in 2006 had potentially injurious infestations (means greater than 799 soybean aphids per plant) that exceed action thresholds for this pest. These results show varying levels of resistance among lines that are homozygous for the Rag1 resistance allele and that protection may be equivocal in years of heavy infestation by soybean aphid. Implications for testing putatively aphid-resistant soybean selections in the field and the potential for field-deployment of aphid-resistant lines are discussed.