The purpose of this paper is to present the study concerning the evaluation of the repair logistics of gas pipeline Urucu-Coari-Manaus (extension of 600 km), that was constructed to operate in the Amazon Brazilian region. Repair logistics is a challenge, regarding specific operation conditions in the jungle, environment and flood variations, difficulty on accessing pipeline right-of-way, difficulty on transportation, etc. Workshops were held, gathering most experienced company personnel from different PETROBRAS sectors (engineering, operation, repair centre, integrity area, Brazilian Army, offshore sector, etc.), in order to evaluate and establish strategies for each identified failure scenario, considering type of repair, logistics, resources and costs. The first step of the study was to incorporate the experience obtained from the engineering team, responsible for the construction of Urucu-Coari-Manaus gas pipeline as they had to face unexpected and adverse conditions. Based on their experience, different pipeline sections were defined, considering specific features, like isolation, flooded areas, river crossings, access limitations, etc. The second step was brain-storming workshops with the purpose of providing the best PETROBRAS evaluation of pipeline sections repair strategies, logistics and resources. Failure frequencies were raised and addressed, as well as variables like: - time for failure detection, for digging, for repair, for resources arrival, considering different logistics and transportation modes (using specific boats, helicopters with special characteristics, such as suitable for long line operations (load line greater than one rotor diameter in length), capable of transporting heavy equipment, etc.). Innovative ways of repair were conceived and proposed to be used. Supply contract conditions for thermo plants, industrial and residential consumers were considered. Finally, a cost/benefit analysis was performed, considering expenses on logistics and resources and benefits associated with avoided losses for each specific failure scenario, in order to provide support for decision making process.