On three occasions, roughly 40 years apart, the mass of the official copies, the national prototypes and the working standards of the BIPM have been compared with the mass of the international prototype. On the last of these occasions (1988-1992), 34 standards from national laboratories were cleaned with solvent and then steam-cleaned with double-distilled water before the definitive mass comparisons.

The results of comparisons between the official copies and the international prototype show some divergence with time. The graph below shows changes of about 5 x 10^–8 in the mass of the standards since their first calibration. All measurements are with respect to the international prototype. For this reason, the mass of the international prototype defines the y-axis of this graph.

There is no reason to suspect that the mass of the international prototype of the kilogram is more stable than its official copies. The question of stability can only be answered definitively by comparison with a fundamental constant of nature (such as the Planck constant or the mass of an atom of ²⁸Si). Such experiments are being carried out in several national laboratories. So far, relative uncertainties of about 10^–7 have been achieved.