In this paper, we describe the preliminary stages of the development of a mathematical model of the evolution of the solution chemistry within a corroding crevice on passive stainless steel. It is based on a formulation by Oldfield and Sutton , but models the physical and chemical processes which determine the crevice solution in a more rigorous manner. The model will eventually be used to assess whether a ‘critical solution composition’, which results in the depassivation of the crevice and the onset of localised corrosion, is attainable for a range of repository conditions, steel types and canister designs etc. We also describe experiments that provide input data in the form of passive currents for this model. Preliminary sensitivity tests with the model have indicated a need for accurate thermodynamic data for the chemical equilibria constants (particularly those for the chromium reactions). These tests also suggest that there are certain critical relationships between various parameters in the system (such as crevice dimensions, the composition of solution outside the crevice and the passive current) that mark different behaviour in the evolution of the solution composition. Further experiments will be performed, as part of this work, both to validate the predictions of the model and to determine whether the predicted compositions of the crevice solutions are sufficiently aggressive to initiate crevice corrosion.