Computerized adaptive testing (CAT) has become increasingly common in large-scale testing programs. This paper considers relevant practical issues that are likely to be faced by the developers and managers of a CAT program. The first cluster of issues is that of item pool development and maintenance. It includes such considerations as item pool specifications, the choice of item response theory model, and other concerns in constructing and choosing test items. The second cluster of items involves administering and scoring the CAT. Proficiency estimation method, test items, item review, and equating CAT scores to paper-and-pencil tests are areas that must be considered. The third cluster involves protecting the integrity of the CAT item pool, considering security and coaching concerns. A fourth cluster includes issues involving examinees. These issues (whether or not to allow item review, how to set time limits, how to address examinee anxiety, test taker motivation, and test equity) are areas that must be explored for fair and useful tests. (SLD)