The numerous reports on lysis of blood (trypomastigote) forms of Trypanosoma cruzi by specific antibodies plus complement have systematically shown that a certain proportion of parasites survives. However, it is not known whether the insensitive organisms represent a subpopulation (or clones) or a certain developmental phase of otherwise morphologically identical parasites. In this work, we established that partial lysis was not due to the use of insufficient amounts of lytic reagents. Thus, supernatants of lytic reaction mixtures killed the same proportion of T. cruzi as previously unused reagents. Moreover, in parallel tests in which the trypomastigote concentration was up to four times greater than that used in standard lysis tests, the percentages of lysis were comparable. Incubation periods as long as 4 h did not increase the extent of lysis beyond the value observed after only 1 h, indicating that the routinely used 1-h incubation was appropriate. The extent of lysis was not increased by additional amounts of antibody, complement, or both. Instead, trypomastigotes surviving immune lysis, washed, and incubated with fresh diluent for 45 to 120 min before being used in new lysis tests did manifest additional sensitivity to immune lysis. Three successive infections in mice with parasites which had survived immune lysis led to the production of trypanosomes that displayed the same level of resistance to immune lysis as the original, untreated parasite population. Of interest, the average parasitemias of these groups of mice did not evidence a tendency to increase, as might have occurred if an immune-lysis-resistant subpopulation had been selected. Since trypomastigotes exhibiting resistance to immune lysis can eventually become sensitive, resistance to immune lysis does not represent an insensitive parasite subpopulation. This resistance appears to be modulated by the presence of the lytic reagents and might involve expression of as yet unidentified surface components playing a role in complement activation.