A method for the determination of total selenium in nutritional supplements and selenised yeast is described. The samples were ashed in nitric acid. Hydrochloric acid was used to prevent precipitation of, in particular, iron salts. After appropriate dilutions, the selenium was determined by Zeeman-effect background corrected graphite furnace atomic absorption spectrometry. A furnace ashing step at 1100 °C was necessary in order to obtain a total recovery of selenium when present in the organic form. Palladium nitrate-magnesium nitrate was used as a matrix modifier. Independent methods were used to determine the content of selenium in a selenised yeast check sample. Accuracy was assured using this sample and by recovery experiments. Between-day random error showed a coefficient of variation of 4.2%. Results from the analysis of eight different commercial supplements were in good agreement with declared contents.