Reference materials do fulfil a very useful function in that they provide means to check, verify and calibrate measurement procedures and instrumentation. However, the material is often “idealized” compared to “real life” samples: it is either (very) pure or does not correspond in concentration level or matrix to the “real life” case. It is shown how this situation could be remedied if “real life” samples with carefully assessed “reference values” could be made available regularly as unknowns to measurement laboratories. Comparison of the latter's measurements with the “reference values” then would provide a picture of the real performance of the particular measurement community in general and of each participating laboratory in particular. The opinion is expressed that isotope-specific methods have now matured to the stage that, if they are correctly applied under rigorous control and based on highly skilled expertise, they can provide such “reference values” against which laboratories can then evaluate their routine performance in regular Interlaboratory Measurement Evaluation Programmes (IMEPs).