A set of methods has been compiled to compare gridded air emission inventories that have been derived, on the same spatial grid, using widely differing techniques. Top-down and bottom-up inventories for Milan, Italy and for the Greater Athens area (GAA), Greece were used to test and apply these methods. The applicability of each method to certain source sectors was assessed by conducting sensitivity analyses. Whilst some of the methods (such as regression calculations or the Moran coefficient) appeared very sensitive to variations of point source emissions, others (e.g. the construction of difference maps) proved more appropriate for characterizing line source differences. Area sources could best be dealt with by using the newly developed acceptability criterion. The development of these tools allows a reproducible comparison of sets of emission inventories and consequently supports developments towards improvement.