When subscores on a test are reported to the test taker, the appropriateness of reporting them depends on whether they provide useful information above what is provided by the total score. Subscores that fail to do so lack adequate psychometric quality and should not be reported. There are several methods for examining the quality of subscores, and in this study seven such methods, four of which are based on classical test theory and three of which are based on item response theory, were reviewed and applied to empirical data. The data consisted of test takers' scores on four test forms--two administrations of a first version of a college admission test and two administrations of a second version--and the analyses were carried out on the subtest and section levels. The two section scores were found to have adequate psychometric quality with all methods used, whereas the results for subtest scores ranged from almost all scores having adequate psychometric quality to none having adequate psychometric quality. The authors recommend using Haberman's method and the related utility index because of their solid theoretical foundation and because of various issues with the other subscore quality methods.