Prior to the landmark discovery in 1960 of ' The Experimental disease that fo llows the injection of human leprosy bacilli into fo otpads of mice,' 1 the only means of searching for drugs active against human disease was to conduct clinical trials. Because clinical improvement of lepromatous patients is both very slow and variable, because the number of AFB (BI) in the skin fa lls extraordinarily slowly despite adequate therapy, and because the viability of solid-staining bacilli (MI) was not appreciated, early short-term clinical trials were difficult to conduct and the results even harder to interpret. Of the earlier studies on dapsone only the study of Lowe2 fo llowed a stable population until bacteriological negativity, finding 32 of 39 (83%) negative at 5 years, 31 of 35 (89%) negative at 6 years, and 34 of 35 (97%) smear-negative at 7 years.The earliest studies on the effect of antimicrobial agents on My cobacterium leprae-infected mice utilized primarily drugs known to be effective against M. tuberculosis. These first studies utilized constant treatment from the time of mouse fo otpad infection, generally with 5 x 103 M. lepraejfootpad, either by incorporation of drug into mouse chow or daily (actually usually five times weekly) intraperitoneal injections. By these means Shepard3 fo und dapsone, clofazimine, isoniazid, para-aminosalicylic acid, streptomycin, and cycloserine active and ethambutal and pyrizinamide inactive. By similar means Gaugas4 in 1967 fo und the same drugs active, and notably, as well, rifampicin and cephaloridine. This method of drug testing, termed the 'continuous method', does not distinguish between purely bacteriostatic and bactericidal activity. Since it has been well established in bacterial endocarditis and a number of other infectious diseases where protective local or systemic host defence mechanisms are inadequate (osteomylitis, meningitis, and Gram negative bacteremia in the neutropenic patient) that bactericidal therapy is crucial to a salutary outcome and that the key to effective short-course chemotherapy for pulmonary tuberculosis is the inclusion of two or more bactericidal agents,5 actual bactericidal activity against M. leprae is likely to be similarly important in the therapy oflepromatous 0305-75 1 8/86/057137S + 12 $0 1 .00