In IMEP-8, two samples of high purity CO(2)(g), with different carbon and oxygen isotope ratios were distributed to 27 participants, originating from 14 countries and from various isotopic measurement domains (geochemistry, atmospheric and food chemistry), but particularly set up for food laboratories. In total 19 laboratories reported results. The outcome of this comparison exercise shows that the laboratories reported carbon and oxygen isotope results in good agreement with the reference values across the domains. The reported results for delta(13)C(VPDB) (carbon) for both materials are within 1 per thousand. However, for the reported results of delta(18)O(VPDB) (oxygen) for both materials the overall spread of the reported results is about 11 per thousand. Within this spread two distinct groups of participants can be identified, where the results within each group vary about 2 per thousand. The latter seems to be caused by calculation errors by participants of the reporting delta(18)O(VPDB) values. As requested, participants also reported the isotope amount ratio for carbon and oxygen in the CO(2) samples. For carbon, all reported results for both materials agree with the isotope ratio value, which can be traced back to the value reported by Craig. For oxygen, all results are in good agreement and deviate by a maximum of 0.5% from the reference values measured at IRMM. Work carried out indicates the carbon isotope ratio, for both samples IMEP-8A and IMEP-8B, differ from those reported by Craig by as much as 1.2%. In the case of oxygen, this deviation is far smaller. Both data sets, i.e. the one realised by Craig and the one realised at IRMM, demonstrate traceability to SI. It is clear that both values significantly disagree.