Tests of Heterodera glycines on differential host plants to determine races were run in Arkansas, Illinois, and North Carolina to check the uniformity of results of the test. Methods used at the three locations varied somewhat. Results indicate that the race test is highly variable. Isolates previously identified as race 1 were identified as race 1 or race 3; those identified as race 2 were identified in these tests as race 2, 4, 9, or 14; those previously identified as race 3 were identified as race 1 or race 3; those identified as race 4 were identified in these tests as race 4 or race 14; those previously identified as race 5 were identified as race 2; and those previously identified as race 6 were identified as race 1, 2, 4, 5, or 6. Part of the variability resulted from the use of differential host plants from different sources and part from nonstandard differential host plants. Other variations may be due to inability to obtain completely uniform inoculum or to recover all nematodes that penetrated.