In order to look for evidence of possible racial bias in six of the Navy tests for the Enlisted Advancement System, statistical data for a group of black personnel were compared with data for a selected matching group of whites. The first step was to compute item difficulty indices (the percent passing each item) for the black group, for the matching white group, and for a separate group of white personnel that were not matched with either group. A correlation was then computed for each test, between the item difficulty indices for the black group and the item difficulty indices for the nonmatched white group. Similarly, a correlation was computed for each test, between the item difficulty indices for the matched white group and the item and the item difficulty indices for the nonmatched white group. This experiment was repeated separately for four different pay grade levels, providing a total of 24 comparisons. In all comparisons, more similarity was found between the two white groups than between the black group and the nonmatched white group, although some differences were small. It was concluded that the tests may have some across-race bias, but no changes in test construction procedures were recommended at the time. (CTM)