Abstract An attempt to separate niobium and tantalum by cupfcrron was only moderately successful at pH 4.5 to 5.5 in the presence of a magnesia mixture as a coagulating agent. A more satisfactory separation of niobium and tantalum from each other, tried out up to ratios of 30:1 and 1.30, is effected with Sn +2 or Sn +4 as a co-precipitating agent under the conditions described niobium can be separated, in the presence of complexone III, from almost all the ions except U, Be, Ti and PO 4 -3 . Iron and other tervalent elements, when present in 100 fold excess with respect to niobium, require double precipitation The method gives highly satisfactory results when applied to the analysis of niobium in niobium-molybdenum stainless steel. The use of titanium as a co-precipitant is less successful than that of tin