ABSTRACT.-Marinho J.B.R., Carvalho A.U., Pierezan F., Keller K.M., Riet-Correa F. Center-West regions and in coastal areas of the Northeast Brazil. Spontaneous poisoning was described in cattle, goats and sheep, with clinical signs evidenced within 24 hours after ingestion of the leaves and death within 48 hours after signs onset. The clinical signs observed in acute poisoning are apathy, anorexia, ruminal arrest, arched back, constipation with feces in small spheres, sometimes covered with mucus and blood streaks, muscle tremors, staggering gait and sometimes sialorrhoea. Neurological signs may be observed, due to interference in the urea cycle due to hepatic insufficiency resulting in hyperammonemia (hepatic encephalopathy). The main pathological finding is centrilobular hepatic necrosis. The toxic principle present in C. axillare was not yet definitively proven, but some authors attribute the toxicity of the plant to the presence of saponins gitogenin and digitogenin. However, it has not been determined whether the saponins present in C. axillare are responsible for the hepatotoxic effect of the plant. Thus, the objective of this work is to determine if the saponins are the compounds responsible for the hepatotoxic effects produced by the ingestion of the leaves of C axillare, using goats as experimental model. For this, the effects of the administration of the leaves were compared with those produced by the saponins isolated from the leaves in goats. Six goats were randomly assigned to three experimental groups that received [1] dry leaves of C. axillare (animals A1 and A2), [2] saponins extract from leaves (animals S1 and S2) or [3] control group (animals C1 and C2). For goats receiving the dry leaves the administered dose of plant was 10g/kg for one animal (A1) and 5g/kg for the other one (A2). For animals receiving the saponins extract, administration was done at a dose equivalent to 20g/kg repeated after 24 hours. The dry leaves administered at a dose of 10g/kg to a goat produced toxic effects, with alterations in biochemistry (indicating hepatic lesion) and histopathology showing centrilobular hepatic necrosis. At the dose of 5 g/kg of dry leaves, clinical signs of poisoning were not observed, but hepatic necrosis was found; after 15 days after the last administration, the hepatic parenchyma of this animal was already normal, with only hemorrhagic areas, demonstrating full regeneration. The administration of extracts of saponins containing gitogenin and digitogenin to goats did not produce significant toxic effects, proving that these compounds are not responsible for intoxication. In addition, goats are a good experimental model for studies of this intoxication.