A set of 35 uranium ore and 10 yellow cake samples, collected worldwide from different mines and production sites, were analyzed
for their impurity spectrum by ICP-MS. Pattern recognition techniques such as cluster analysis were applied to the data set
in order to characterize samples with relation to their geographical origin. The results obtained show a clear relationship
between samples taken from the same geological origin and constitute a satisfactory fingerprint for establishing the origin
of the material. In addition to the impurity data, data on the isotopic composition of radiogenic lead is used to resolve
ambiguity when impurity cluster analysis fails to deliver unambiguous origin data.