Recently reticulocyte count testing has markedly decreased in Japan. We studied the current status of the reduction of reimbursement for reticulocyte count testing and the cost-effectiveness of the test in an academic hospital. The orders of complete blood counts(CBC) accompanied by reticulocyte count testing in Tokai University Hospital declined from 58.9 to 13.3% during the last decade. Among a total of 15,538 test orders for reticulocyte count in outpatients, full reimbursement was refused in 123(0.79%) during a one year period from April 1999 to March 2000. The cost-effectiveness of reticulocyte count testing was evaluated in 719 new outpatients who visited Tokai University Hospital during this period. A useful result(UR) of testing was defined as that which contributed to a change in a physician's diagnosis- or decision-making, relating to a "tentative initial diagnosis" obtained from patient history and physical examination alone. Cost-effectiveness for either CBC alone or in combination with reticulocyte count testing was evaluated by calculating clinical effectiveness(the number of URs yielded by a test per patient), economic efficiency(cost required for a test per UR generated), and the cost-effectiveness values(cost required for additional test(s) per additional UR generated). There was a wide disparity in the clinical effectiveness of reticulocyte count twisting among the hospital departments, ranging from 0.53 UR/patient(Gastroenterology-2) to 0(Oral Surgery, Neurology, Urology). Poor clinical effectiveness shown in particular hospital departments endorsed the increasing percentages in reimbursement rejection by the payer of health insurance. The test should be ordered more carefully in consideration of reimbursement and in regard to its cost-effectiveness among different patients groups.