Differential reduced minus oxidized (Red-Ox) or reduced. CO minus reduced (Red. CO-Red) spectra of Trypanosoma cruzi metacyclic trypomastigotes (Tulahuen strain), revealed the presence of cytochromes aa3, b, c5 5 8, o and possibly d (Fig. 1). Mitochondrial membranes from the epimastigote form of the parasite showed similar cytochrome spectra (Fig. 2A). Extraction of the mitochondrial membranes with guanidine and cholate, and spectroscopy of the cytochrome o -cyanide derivative proved that this cytochrome is an integral constituent of the mitochondrial membranes (Figs. 2B and 4, and Table 1). Contamination of the investigated samples with hemoglobin, peroxidase, catalase, cytochrome P-420, cytochrome P-450 or culture medium hemoproteins was ruled out by filtering the cells and mitochondrial membranes on Sephadex G-50, or by differential spectroscopy of pigments, or by differential centrifugations of samples. Investigation of pyridine-hemochromes revealed hemes A, B, C and D (Fig. 3), thus confirming the presence of the postulated cytochromes. Comparative spectroscopy of a series of T. cruzi stocks (including Tulahuen and Y strains), many of them obtained from acute or chronic forms of Chagas disease, revealed significant variability in the cytochrome content (Table 2). Taking cytochromes o and b as standard for comparison, the epimastigotes samples could be grouped as follows (in parenthesis number of passages through the culture medium): 1) stocks with a relatively high content of cytochromes b and o, prevailing the former (stocks Y (116), RA (114), AF, FN, TN and MG (14 y 16); 2) stocks with a relatively low content of both cytochromes: Y (119), AWP and UP; 3) stocks with a low content of cytochrome b, without cytochrome o: CA-I and CA-I (V); 4) stocks without cytochromes: Y(117 and 118) and RA(113). In some strains (e.g. Y and RA), significant variation of cytochrome content in different stocks of the same isolate was observed (Table 2), but the Tulahuen strain proved to be less variable. Comparison of cytochrome distribution and other properties of parasites, namely, lethality for mice and morphology, did not allow to establish positive correlations.