This paper outlines the procedures used to implement and test the CARL version C100 circulation system at the University of Colorado's Boulder Campus. Intended to serve as a guide to other libraries implementing new circulation software, the paper describes a broad approach to testing such software, describes procedures that speed and organize the process, and suggests staffing levels and equipment to have available. The focus is on describing methods to test the accuracy of data transfer, determine the functionality of the features, and to confirm that the local parameters interact properly with all circulation functions. The following steps are covered: (1) test preparations, such as involving technical staff and organizing work area and materials; (2) documenting the transfer of data from the old system to CARL by preparing forms for checking the accuracy of the transfer and providing a means, such as fax or email, to document discussions with CARL about problems identified during the test; (3) implementing test procedures to check the accuracy of the transfer of data on materials, patron records, and transactions; (4) testing to make sure functions and features, such as In-transit, Added Borrower, and Renew-All, perform as they are supposed to; (5) stress testing the system by such means as adding 500 books to a patron's account and renewing them all; (6) alert CARL that the system is ready to go live after identified problems are cleared up; and (7) continuing to retest critical elements such as parameters and functions after the system goes live. The prospective user of these guidelines is warned that although testing is time and labor intensive, it is essential; even though others have implemented the system, there may still be bugs that have not been resolved. (KRN)