The status of testing has dramatically improved over the past 15 years, with a subset of the plethora of tests becoming more used and more dependable. There is still a tendency, however, for each set of researchers to create yet another test, and this author is not exempt from this criticism. The justification for another test is that Song and Hattie test was created to specifically test theories of the underlying structure of self-concept and it is not suggested that it be used for diagnosis, placement, or remediation of individuals. It has been used for assessing the effect of some programs and it particularly intended to assess relationships between the various dimensions of self-concept. The more widely used scales are subject to more and more criticism, and researchers can begin to feel more confident in using these scales to assess self-concept. Although the estimates of reliability are typically very high, there is still a problem with validity. A major concern is that there are so few validity studies. The few that have been published relate to concurrent validity, and there is an absence of construct validity studies. The Piers-Harris has been extensively used with primary school children and as a consequence, there is much psychometric data available. The consistency and stability estimates of reliability are very reasonable and it has high correlations with other reliable self-concept tests. The evidence for validity is promising and the analysis of the dimensionality provides support for the behavior, intellectual and school, and physical appearance and attributes scales. There is some support for the anxiety and popularity scales, although it is suggested that the happiness and satisfaction scales be separated. Underlying the Piers-Harris scales are two second-order factors tentatively identified as academic and nonacademic self-concept. The Coopersmith has also been extensively used in many research studies, although it has yielded less impressive psychometric evidence than Piers-Harris. Although estimates of consistency and stability are satisfactory for most scales, the test needs more validity evidence to be convincing. The dimensionality of the Coopersmith is far from clear anr there is support only for the general and home scales. The most promising new tests of self-concept are the Marsh SDQs. The various versions are becoming more widely used, it is based on the Shavelson model, and there is much evidence accumulating from various research teams to support its usefulness and to provide information on its psychometric properties. A major disappointment has been the lack of measures to differentiate self-concept and self-esteem. As it is possible to philosophically distinguish between the two constructs, this is an area where more research is critically needed.