The different modes of transport used in the agroethanol industry and their energy efficiencies have been studied. Their specific consumption of fuels t(trans) in MJ (t(load) km)(-1) is assessed from raw data and from friction force laws. t(trans) depends on the mode characteristics, fuel/engine performance, velocity, geometry, total mass, actual load... Lack of precision on them increases the uncertainty on t(trans) (variation by a factor up to 8 for pipeline depending on the flow velocity). From t(trans) is deduced the consumption of the mode in the industry R(trans) in) for 100 J of the energy content of ethanol E(etoh) produced from the load. R(trans) takes also into account the distance of shipment d and the weight of the load in E(etoh), W(load). Trucks, t(trans) from 7 to 1.4 MJ(t(load) . km)(-1,) can present the best R(trans), lower than 0.5 J for 100 J of ethanol, because of trips over small d (less than 100 km) and of low w(load) (less than 0.04 t(load).GJ(etoh)(-1) for farm inputs and ethanol). R(trans) of the plant transport to the factory by trucks ranges to 3,1 due to larger w(load) (up to 0.56 t(load).GJ(etoh)(-1) for sugar cane). Large part of the ethanol is moved from the factory to the local storages over 1000 km more or less depending on the proximity of consumption centers. Efficient modes such as pipeline and sea ships, ttrans as low as 0.05 mJ (t(load).km)(-1) when optimized, can compensate for these distances with R(trans) around 1 J. R(trans) to export ethanol from Brazil to France would represent less than 5 J, much lower than the difference of consumptions R between sugar cane and sugar beet based ethanol productions. (C) 2011 Elsevier Ltd. All rights reserved.