The distribution of some key impurities in high-purity copper cathode was found to be significantly biased towards edges of the cathodes. It is therefore necessary that the laboratory sampling technique applied to primary sampled cathodes proportionately contain edge materials. Even with high-purity cathode lots, variation within and between cathodes in the lot is large; thus sampling accuracy requires that both primary sampling frequency and the laboratory sampled portion of whole cathodes be relatively high compared with most other metallic materials. There appears to be no low-frequency simple sampling procedure that is a reliable in-plant indicator of general quality.