Reproducibility of overbank sediment sampling was tested in twenty-nine floodplains in Europe, ten in Greece and nineteen in Norway, by the collection of duplicate pairs of samples. Distances between duplicate sites in Greece were 60 to 100 m, and in Norway 100 to 200 m. In Norway the same nineteen floodplains were sampled by a second team for the purpose of investigating differences in sampling variability and technique. Total element contents were determined in all samples. Paired samples were compared by calculating Spearman's rank correlation coefficient on the raw analytical data, and one-way analysis of variance on the log-transformed data. Pairs of overbank sediment samples collected from different floodplains by the Hellenic team and the first Norwegian team showed high rank correlations and low within-basin variability (sampling and analytical variance). Statistical results of the second Norwegian team were comparatively poorer; both Spearman's rank correlation coefficient and one-way analysis of variance, showed very low positive to negative correlations and high within-basin variation, suggesting a non-uniform distribution of elements in the Norwegian overbank sediment sequences and differences in the sampling technique of the two teams. Nevertheless, careful location of sample sites, as has been done by the Hellenic and the first Norwegian teams, reduces considerably the sampling variability, and the overall sampling reproducibility for most elements is very good for distances up to 100 m in Greece and 200 m in Norway, provided correlated overbank sediment sequences are sampled. The implication of this study for multinational regional geochemical mapping is that overbank sediment sampling must be carried out by well-trained professional teams of exploration geochemists, and where possible by one sampling team for the whole country.