A case study of the Florida Teacher Certification Examination (FTCE) program was described to assist others launching the development of large scale item banks. FTCE has four subtests: Mathematics, Reading, Writing, and Professional Education. Rasch calibrated item banks have been developed for all subtests except Writing. The methods used to evaluate the stability of the score scale as it relates to a cut score based upon a logit value are explained. Cut off levels were established in terms of a logit ability scale based on field test data. Linking items were chosen so that difficulty levels were centered at the cut score. Item selection was based upon closeness of fit to the Rasch Model and item difficulty. The linking items for the Professional Education subtest were of primary concern. Results showed no bias in the test due to differences in curriculum. (DWH)