Abbey's select laboratories method has been applied to the reported data for major and minor elements in the proposed GIT‐IWG reference sample IF‐G. The final values obtained by this method have been compared with the recommended values published by Govindaraju in 1984. Slightly different percentages are suggested for MgO, K2O and Fe2O3T. The importance of applying the Abbey method by participants in co‐operative studies is emphasized. Such an application gives a better understanding of data treatment and of the degree of reliability of published final values; furthermore the objective score obtained for laboratory quality will stimulate participating laboratories to improve their analytical methods and procedures. As an illustration some improvements in the laboratory of the authors are ment