The behavior of a Colombian strain of Trypanosoma rangeli was investigated in Rhodnius prolixus, its classical biological vector, and compared with that in 8 other species of Triatominae of the genera Rhodnius and Triatoma. Infections were accomplished by feeding the bugs on infected mice or on suspensions of culture forms, or by inoculation of culture into the hemocoel. The 536 insects used were examined repeatedly and the results were assessed statistically. Intestinal and spontaneous hemolymph infections were observed in all species; invasion of the salivary glands, demonstrated at dissection, by bite transmission, or both, was observed in Rhodnius prolixus (18%), R. neglectus (38%), Triatoma patagonica (11%). T. protracta (7%), and T. infestans (3%), the rates in R. neglectus and T. infestans being statistically higher and lower, respectively, than those in R. prolixus. T. rubrovaria had the highest frequency of hemolymph invasion but the glands were never infected. Parasites were seen in the hemolymph of a high proportion of the insects up to 143 days. Loss of gland infection was observed in 8% of R. prolixus and 20% of R. neglectus, but not in Triatoma spp. The absence of gland infection in T. lecticularius and T. phyllosoma pallidipennis harboring trypanosomes in the gut and hemolymph and the complete absence of infection in T. maculata need reassessment, since the trypanosomes to which these species were exposed were of low infectivity at the time of the experiments. It seems that frequent invasion and high reproduction rate of Trypanosoma rangeli in the hemolymph do not necessarily indicate the capacity of a species to act as vector. To the previously known natural Vectors (R. prolixus, R. pallescens, R. ecuadoriensis and possibly Triatoma dimidiata capitata), must now be added 4 experimental ones: R. neglectus, T. patagonica, T. protracta and T. infestans. Trypanosoma rangeli has not been reported from large areas where these species occur (parts of Brazil, Mexico, the United States and southern South America), perhaps because in the absence of the best known vector, R. prolixus, adequate surveys have not been carried out.