Chagas disease caused by a protozoan called Trypanosoma cruzi is a neglected tropical disease and a leading cause of heart failure in Latin America where it is endemic. Due to migration of asymptomatic infected population, it is now also present in North America, Europe, Japan and Australia. Drugs currently available to treat Chagas disease are limited to benznidazole and nifurtimox, both with severe side effects, and are more efficient when administered early in the course of the infection. Thus, there is a need for the discovery of alternative and more efficient drugs. In the identification process of potential drug targets, metabolic genes are good candidates, since they are critical for cellular growth and survival [1]. In addition, genome-scale metabolic models (GEMs) [31] have been developed to accurately predict metabolic capabilities from genome sequences. This work developed an extended GEM, hereafter referred to as iIS312, of the published and validated T. cruzi CL Brener core metabolism model iSR215 [35]. The life cycle of the parasite is divided in three distinct stages: epimastigotes (replicative and insect-specific), trypomastigotes or metacyclic epimastigote (infective and non-replicative) and amastigotes (replicative and intracellular in humans). From iIS312, we built three stagespecific models using transcriptomics data [3] integration. Such models for other organisms (e.g. Plasmodium falciparum) have provided valuable insight of stage-specific changes in metabolism [1]. The stage-specific models were able to predict metabolic differences among the three T. cruzi stages, including essential genes and reactions, and flux distribution. The stage-specific models showed that epimastigotes present the most active metabolism among the stages, with least number of deactivated reactions. The trypomastigote model predicted the non-essentiality of pathways responsible for nucleic acid synthesis as this stage is non-replicative, and presented full metabolic activity in TCA for energy yield to move around. The amastigote model presented low activity in Glycolysis, as observed in previous studies, suggesting that this pathway is not essential in amastigotes. It also indicated that amastigotes take a shortcut in TCA, increasing the metabolic flux in the malic enzyme, as described in previous studies. Moreover, the gene essentiality predictions suggests potential drug targets, among which some have been proven lethal previously, including glutamate dehydrogenase [12], glucokinase and hexokinase [38]. Our results represents potentially a step forward towards the improvement of Chagas disease treatment. To our knowledge, these stage-specific models are the first GEM built for the stages amastigote and trypomastigote. This work is also the first to present an in silico GEM comparison among different stages in T. cruzi life-cycle.