This paper comments on the contributions to this special issue on item banking. An historical framework for viewing the papers is provided by brief reviews of the literature in the areas of item response theory, item banking, and computerized testing. In general, the eight papers are viewed as contributing valuable technical knowledge for implementing testing programs with the aid of item banks.As noted by van der Linden (1986), measurement models and testing practices have undergone some major changes in the last 20 years. One important change is the shift from the nearly exclusive use of classical test models and standardized tests to the use of newer test models derived from item response theory (iRT) and customized tests (i.e., tests that have been constructed to match user needs).Test administration procedures have also changed: Group-administered paper-and-pencil tests are sometimes replaced by tests administered adaptively at computer terminals. There has also been a major shift in emphasis from norm-referenced to criterion-referenced test score interpretation and a concomitant desire in the testing field to tie testing and instruction more closely together.There are signs of change throughout the testing industry. Large-scale testing programs such as Educational Testing Service's Scholastic Aptitude Tests use IRT models to equate test forms. Major test publishers such as CTB/McGraw-Hill and The Psychological Corporation are using IRT models to develop tests and to equate test scores. They are also using IRT models to help produce customized standardized achievement tests. Departments of education in states such as California and Maryland are using IRT models in their state assessment programs in reporting scores and addressing item bias. The College Board's popular Degrees of Reading Power, which links test results to reading instructional levels, uses IRT as the measurement model to accomplish its goals. School districts such as Portland, Oregon and Los Angeles, California are using item banks and achievement scales which are based on IRT models to deliver tests and score information to schools and homes. The National Assessment of Educational Progress, a major federal testing program aimed at helping policymakers monitor the quality of education in the LT. S . , is using IRT models in scaling scores and reporting information. Computerized adaptive testing is now in use or under serious consideration by the ~J.S. Armed Services and several of the large medical and allied health organizations.