Several methods have been used to isolate bloodstream forms (trypomastigotes) of trypanosomes from other blood components. Four of them have been widely applied: (1) defibrination of the blood by incubation in a medium containing calcium chloride and shaking with glass beads followed by passage of the parasites through a sintered glass funnel (Fulton and Spooner, 1957, Ann. Trop. Med. Parasitol. 51: 417-421); (2) defribination followed by low speed centrifugation in Krebs-buffered-glucose-saline layered on density gradients of sucrose (Williams and Cover, 1966, Trans. R. Soc. Trop. Med. Hyg. 60: 426-427; Jaffe et al., 1969, Exp. Parasitol. 25: 311-318); (3) separation using a gradient of N-methyl-3-5-diacetamide-2-4-6-triiodobenzoate and Ficoll (Budzko and Kierszenbaum, 1974, J. Parasitol. 60: 1037-1038); (4) separation using a DEAE-cellulose column (Lanham, 1968, Nature 218: 1273-1274; Lanham and Godfrey, 1970, Exp. Parasitol. 28: 521-534; Gutteridge et al., 1978, Parasitology 76: 159-176; Villalta and Leon, 1979, J. Parasitol. 65: 188-189). Electron microscopic observations made on Trypanosoma cruzi isolated by the first three methods indicate variable contamination with platelets (unpubl. observ.). The fourth method, which is being used routinely for isolation of Trypanosoma brucei, presents two problems. First, erythrocytes and bloodstream forms of T. cruzi have a similar surface charge (de Souza et al., 1977, J. Protozool. 24: 411-415). This problem was solved by a preliminary separation of the trypanosomes from the erythrocytes, using low speed centrifugation leaving the DEAE-cellulose column for separation of the trypanosomes from other blood components (Gutteridge et al., 1978, loc. cit.; Villalta and Leon, 1979, loc. cit.). Second, several lines of evidence indicate that passage of trypomastigotes of T. cruzi through DEAE-cellulose column may alter some properties of the parasite such as the transport of amino acids (Goldberg et al., 1976, J. Protozool. 23: 79-186), the surface charge (de Souza et al., 1977, loc. cit.), and infectivity to mice (Villalta and Leon, 1979, loc. cit.). In view of the problems pointed out above we decided to develop a method for separation of bloodstream trypomastigotes of T. cruzi by using a gradient of Metrizamide [2-(3acetoamido-5-N-methylacetamido-2-4-6-triiodobenzamido)-2-deoxy-D-glucose]. This substance has a molecular weight of 789, a density of 2.17 g/cm3, gives solutions of low viscosity and osmolality, dissolves readily in water or dilute solutions of salts, has no ionizable groups, and can be stored indefinitely at -20 C (Rickwood and Bimie, 1975, FEBS Lett. 50: 102-110). It has been applied to isolate several types of cells and subcellular particles. Blood (about 30 ml) obtained by bleeding mice through the retroorbital sinus (1-4.107 trypomastigotes/ml) was collected with medium 199 plus 3.8% sodium citrate as anticoagulant and centrifuged at 500 g for 5 min to pellet most of the erythrocytes, and left at 37 C for 30 min so that some pelleted trypomastigotes could move to the supernatant. The supernatant was centrifuged at 2,800 g for 30 min after which the pellet was resuspended in 3 ml of medium 199 plus 10% fetal bovine serum and left at 37 C for 5 min for aggregation of the platelets. In some experiments, adenosine diphosphate (final concentration of 1 mg/ml) was added slowly with occasional shaking in order to improve agglutination of the platelets. However, this step can be omitted. Pressing the platelet aggregate gently with a Pasteur pipette improved the liberation of some trypanosomes. The suspension containing the parasites plus aggregated platelets and other blood cells was layered on the top