Values of repeatability (r) and reproducibility (R) are parameters of an analytical method. They are related to the maximum difference between two results of analysis (permissible range) and can be obtained by interlaboratorial comparisons. They represent great importance for the quality control of the results and decisions about the fitness-for-purpose of the method. Water determination in fuel oil samples were performed by 16 laboratories, using the Karl Fischer reaction, during the period from June 1997 to December 1999. Values of r and R were presented. Discussion is made in terms of the main factors that would be contribution to analytical method variabilities and uncertainties of final results.