Over the past decade the clinical results of kidney transplantation have improved substantially, with much of the benefit being attributed to the introduction in late 1983 of the immunosuppressive drug cyclosporine. To assess the effect of cyclosporine on the use of hospital services, we studied 702 patients who received kidney transplants at the University of California, San Francisco, between July 1982 and June 1986. All services were priced in constant 1985 dollars, and multiple regression analysis was used to adjust for changing patient and hospital characteristics. The introduction of cyclosporine for patients receiving kidneys from cadavers was associated with a significantly shorter adjusted mean postoperative stay (26.4 days as compared with 37.0 for patients not taking the drug; P less than 0.0001) and lower adjusted mean hospital charges ($28,649 as compared with $37,895; P less than 0.0001), although cyclosporine was not associated with changes in the use of services by patients who received transplants from living related donors. Cyclosporine was also associated with a reduction in the use of certain ancillary services, such as laboratory tests and radiographic procedures. In patients without diabetes who received cadaver kidneys, a sequential cyclosporine regimen (in which a combination of antilymphoblast globulin, prednisone, and azathioprine was given before cyclosporine) reduced the use of hospital services even more than did a cyclosporine regimen in which the combination was not given. The results suggest that new medications, such as cyclosporine, that reduce the frequency of complications and improve outcomes may also reduce the use of hospital resources.