The "Griggs vs. Duke Power" decision of the Supreme Court specifies that job relatedness is the only lawful reason for using tests for the purpose of selection when their use results in the disproportionate exclusion of minority group members. One of the main implications of the "Griggs" decision for the test producer and user alike is a renewed emphasis on sound practices of employee selection. The producer of employment tests must put more effort into job analysis activities, criterion development activities, and criterion related validity studies. Another general concern for the test publisher is about certain characteristics of the test or its administration that are irrelevant to the specific abilities being measured but which depress scores for members of one group. Another consideration is that of differential validity studies. There is more than one reasonable definition of test fairness, and these definitions are in conflict; thus, competing values must be weighed. The test maker cannot resolve the problem of competing values, but he can provide the test user with information that will make explicit the value implications of various uses of tests. Two figures illustrate Thorndike's alternate definition of culture fairness. (DB)