An interlaboratory cooperative study was conducted to determine total selenium (Se) in two lots of a commercially-available (Seronorm) lyophilized human blood serum. The main objective was to arrive at consensus concentration values in order to establish control sera for chemical analysis. A secondary goal was to assess the performance of different analytical methods as applied to the analysis of serum specimens with normal concentrations of Se, i. e. about 100 mu g/L. Twenty seven laboratories worldwide participated in the trial using six inherently different analytical methods: (1) fluorometry, (2) electrothermal atomization atomic absorption spectrometry with or without sample pretreatment, (3) neutron activation analysis (instrumental or with radiochemical separation), (4) hydride generation atomic absorption spectrometry, (5) isotope dilution mass spectrometry, and (6) X-ray fluorescence spectrometry. For this population of laboratories, methods (2) and (4) exhibited the largest systematic differences in concentration.