The European Standards Committee on Oxidative DNA Damage (ESCODD) was set up in 1997 to resolve methodological problems and to reach agreement on the basal level of 8-oxo-2 0 -deoxyguanosine (8-oxodG) in biological samples. In the present ESCODD trial 6 samples of 8-oxodG-containing oligonucleotides with different ratios of 8-oxodG/2 0 -deoxyadenosine (dAdo) were sent to 25 labora-tories throughout Europe. The methods used were HPLC with electrochemical detection (amperometric or coulometric), GC– MS or LC–MS–MS. The LC–MS–MS and the coulometric HPLC analyses gave 8-oxodG concentrations within 53 and 73% of expected values, respectively, whereas the ampero-metric HPLC and GC–MS consistently overestimated the 8-oxodG concentration by several fold. As the oligonucleotides contained no 2 0 -deoxyguanosine (dGuo), this was not due to artificial oxidation. On the contrary, in most cases the concentrations of dAdo and thymidine (dThd), used as estimates for non-oxidised DNA bases were underestimated, though a few laboratories overestimated the lowest concen-tration samples containing 8 and 20 mM, respectively. In one-third of the reported results, the ratio of 8-oxodG/10 5 dAdo was within 25% of the calculated value in the oligonucleotide samples and in half of the results the coefficient of variation in duplicate samples was less than 10%. The coefficients of variation were higher for the dAdo concentrations than for 8-oxodG. Our findings strongly indicate that careful quality control must be applied to the analytical procedures for 8-oxodG and very importantly also to the procedures for non-modified 2 0 -deoxyribonucleosides. We recommend the use of synthetic oligonucleotides for this purpose.