To complement the many findings already reported, RMP serum levels were determined in 60 patients. Determinations were done by the microbiological method of agar diffusion (test organism, Sarcina lutea); and results were assessed by means of standard curves. The patients were receiving long-term therapy with RMP in combination with other tuberculostatic agents. Serum levels were determined before, and 90 min and 3 h after 600 mg of RMP were given on an empty stomach. The absolute serum concentrations measured, the times when peak levels were reached, and the considerable individual variations in some cases, agreed largely with the results of other investigations using comparable methods. Long-term administration of RMP and the combination with other tuberculostatic agents did not affect the serum levels to any notable extent, though the group of patients given RMP therapy for more than 6 months did show definitely the highest serum concentrations. On the other hand it should be stressed that a single total dose every 24 hours is only the best method in the treatment of tuberculosis, and not of other bacterial infections. With the latter, the dose should be preferably divided into two per day, because nonspecific organisms causing infections produce new generations at about 4 times the rate shown by Mycobacterium tuberculosis. Experimental studies seem to indicate that intermittent therapy with high doses promises much for the treatment of tuberculosis, but this still requires further clinical investigation.