Data uncertainty issues have constrained the widespread acceptance of life-cycle analysis (LCA) and related methods. This is particularly important in the LCA of fuels due to the wide range of available feedstocks and processing options. Despite recent attempts at standardization, there remain persistent doubts about the general validity of LCA results, often due to uncertainties about data quality. This paper demonstrates the application of possibility theory as a tool for handling life-cycle inventory data imprecision for the case of the net energy balance of coconut methyl ester (CME) as a biodiesel transport fuel. Results derived using a possibililistic computation are contrasted with those arrived at by probabilistic (Monte Carlo) simulation. The two approaches yield comparable results but possibilistic modelling offers significant advantages with respect to computational efficiency. The net energy balance of CME is estimated to be approximately 36 MJ kg(-1), significantly higher than the 28 MJ kg(-1) net energy typical of rapeseed oil methyl ester (RME) relevant to the U.K. Copyright (C) 2002 John Wiley Sons, Ltd.