Abstract
After a brief introduction indicating the principal function of reference materials (RMs) in ensuring that the analytical measurement programme is performing reliably, the availability, different types and sources of information on RMs are described. Next, the correspondence of available RMs to present sample demands is discussed, particularly with regard to the adequacy of matrices and analytes, the availability of RMs for ‘difficult’ trace elements, and the preparation and certification of speciation RMs. The role of RMs in quality control is indicated, and their usage and certain problems which can thereby arise, including total dissolution, homogeneity and their representativeness in relation to samples, are discussed. The application of truly representative RMs in low-level trace element work, as a means of quality control of sampling and storage procedures, is mentioned. Strategies open to the analyst in the total absence of certified RMs are presented, and the importance of publication of independent results by alternative methods for such materials, as a means of establishing consensus values, is stressed. In the situation where results for a particular analyte obtained by different groups for the same (but not identical) biomedical or environmental system differ markedly, a progressive multi-stage exchange scheme is proposed, which is designed to reveal whether the differences arise from analytical factors, sampling effects or real system differences.