The key to any selection instrument is its validity, i.e., how well the instrument measures the person for the job. The specifics of establishing test validity are contained in the Uniform Guidelines for Employee Selection, which have been adopted by a number of Federal agencies. Three types of validity are recognized: content, criterion-related, and construct. Content validity implies that the test instrument adequately samples and represents the job domain; criterion-related validity connotes that the test either predicts future or current job behavior; and construct validity entails a statistical analysis of test data to establish the data's purity and consistency of measurement. Even though a test may prove to be valid for the selection of police officer candidates, the selection of candidates based on test scores can be done in a number of ways to ensure that an appropriate number of minority candidates are selected. For example, candidates may be selected by racial representation based on a pass-fail rather than a rank order standard. Also, lists of white and minority candidates may be separately prepared, with the highest scorers on each list being selected. 5 tables, 34 footnotes.