The reversals of Mitsuda's reactions induced by BCG have been objected to based on the possiblem interference of other determination causes of the phenomenon: tuberculous primo-infections, communicants of unsuspected leprosy, revearsals due to other causes, such as anti-diphteric and anti-tetanic vaccination, etc. In order to study the problem, we have used Rhesus monkeys (Macaca mulatta), which were reared in isolation, in an attempt to avoid the referred to interferences. Prior to the experiments, all animals were tested and found negative to radiograph, tuberculin and lepromin tests and were then submitted to the application of BCG vaccine (from 1 to 3 days old), in different doses and by different via. At different times, after the application of BCG, they were again submitted to the radiographic, tuberculin and lepromin tests. In the tables I to IV the experiences were summarised. From the experiments, the following conclusions were reached: 1 - From 12 Rhesus that received BCG 11 showed reversals of the Mitsuda reaction (91.7%). 2 - These reverseals took place both in tests effected shortly after BCG (from 6 days to 2 months), and tests effected much later (from 7 to 12 months after BCG). 3 - Some differences were found in the results, according to the dosis and the application via of the BCG. a) - The testicular and peritonela via (0,02g) were the only that determined strong positive Mitsuda's reactions (+++). b) - By oral via, animals that received high dosis (0.6g and 1.2 g), there resulted uniform and regular reversals, even though of low intensity (+); but from those who got small doses (0.2 g.) one showed no reversals in all tests, and the other presented reversals in the 2nd and 3rd tests only, also with low positivity (+). 4) In the 2nd and 3rd Mitsuda's reactions in the same animals, positivity was always precocious (generally within 48 hours), one getting the impression that there occurs a sensibilization of the animal body by the antigen with the repetition of the tests, even though the intensity of the reaction always remains the same. This precocious reaction (Fernandez type) occurs both shortly and long time after the application of the BCG. Its precocity depends not of the antigen only because the first Mitsuda's reaction after the BCG application occurs after some time and seems not influenced by the control lepromin test effected on the Rhesus before the BCG. 5) On the control group, the animals which received a.a.f. bacilli suspensions (Mycobacterium sp.; M. avium, and M. smegmatis), did not show reverseals of the Mitsuda's reaction. Two Rhesus, however, which received dead BCG (120ºC autoclave 1 hour), one intradermically (0.006 g) and the other orally (1.2 g), did both present reversals of the Mitsuda's reaction, with weak positivity (+). In all animals of the control-group, the allergic reactions were found negative. 6) Strong local inflammatory reactions were observed in the Rhesus that had received living BCG by intradermal via, and in the one submitted to multipunctures, there occurred the formation of a large caseous abcess. 7) The allergic tuberculinic and infratuberculinic reactions appeared dissociated from the Mitsuda's reactions: sometimes they are more precocious, occurring before of the lepromin test; on other occasions they disappear, when the Mitsuda's reactions still persist; and finally, they may be absent, when the latter occur, especially after the oral application of the BCG. 8) In Rhesus which received BCG by testicular and peritonela via, in the infratuberculinic test (0.1 ml of total BCG extract), besides the classic answer, which occurs between 48 and 96 hours, one could observe a delayed answer (15 to 20 days), represented by a non-erythematous nodule, which persists for 11-14 days.