An important requirement in every geological quality control program is that the identity and grades of the control samples (duplicates, certified reference materials (CRMs), and blanks) must remain blind' to the analytical laboratory._x000D_
Once the entity to be controlled is aware of relevant details of the control process, such as the identity of the control samples, the objectivity may be lost and the control may no longer be completely valid._x000D_
Whereas the blind insertion of field duplicates and coarse blanks can be easily conducted at site, simultaneously with the sampling process, the rest of the control samples should be inserted into the batches during or after sample preparation and prior to assaying, in order to ensure a proper assessment of precision, accuracy and assay contamination, independent of the internal quality control of the laboratory._x000D_
Such a procedure raises the concern that blind insertion of crushed or pulverised material in sample batches prior to assaying could represent a breach in the chain of custody principle, as samples would have to be temporarily returned to the custody of a non-laboratory person in charge of the insertions._x000D_
In this seemingly conflict scenario, what should we give up? Do we maintain the strict chain of custody, knowing that the laboratory will be aware of the identity of some of the control samples, or that some duplicates will not be true same-batch duplicates? Do we ignore chain of custody issues to maintain the anonymity of the control samples and insert blind same-batch duplicates, CRMs and blanks? Is it possible to do blind insertions without breaching the chain of custody? Current NI 43-101 and JORC requirements, as well as real-world QA/QC and sample security practices, are discussed in light of AMEC's approach to the solution of this apparent dilemma.