After emphasizing the needs for standardization of processes for the clinicai trial of drugs in Cragas' disease, the Authors report a method based in the principie that only patients with "high parasitemia" should be selected for therapeutic trials. When patients with chronic Chagas' disease are monthly submitted to xenodiagnosis, the results are widely variable. While some patients show all the xenos to be positive (100%), others never have a positive xenodiagnosis (0%), even if the period of observation goes for three year (table III). Between these extremes a regular gradation of positivity is observed. It is necessary to establish, before treatment, the parasitoloficai behaviour of the patient as xenodiagnosis is concerned, and only those patients with a positivity of the xenodiagnosis of 60 per cent or higher should be selected for the study. The rules for the trial are presented, in relation to its three phases: pre-therapy, during the therapy (with the patients in the Hospital) and post-therapy. Three cases of a trial, still under way, with a derivative of nitrofurfurilidene (Bayer 2502) are shown, to exemplify the value of xenodiagnosis and sorological tests in the evaluation of efficacy of the drug (tables IV, V and VI). Finally the Authors discuss and define the criteria of cure of human Chagas' disease