Detailed mineralogical investigation of a tailing sample from the Juno plant circuit at Tennant Creek was carried out to determine the nature of the unrecovered bismuth and whether any possibility existed for improving bismuth recovery. The investigation showed that the Ibismuth was present in a complex mixture of Bi-Pb-Ou-Se-S minerals (termed "bismuthinite"), of which about 74 per cent was locked (mainly with magnetite) and 26 per cent was liberated (18 per cent in the minus 7 m slime). The results suggested that attention should be directed to both finer grinding and improved flotation of slime "bismuthinite", and testwork carried out by Peko Mines N.L. on the basis of these recommendations led to a significant improvement in bismuth recovery.Detailed mineralogical investigation of two Juno concentrates was carried out to determine the nature of the selenium present and whether any possibility existed for improving the selenium grade. The investigation indicated that virtqally all the bismuth, selenium, and lead was present in "bismuthinite", which showed a composition range of 32-82 per cent Bi, 0-42 per cent Pb, 0-24 per cent Se, and 5-18 per cent S. Differences in the Bi/Se ratios of the two concentrates could be attributed to different average Bi/Se ratios of the constituent "bismuthinite", and the relatively low selenium grade of the "bismuthinite" precluded the possibility of obtaining a high grade selenium concentrate with high selenium recovery.However, differential flotation and/or low intensity magnetic separation to remove chakopyrite, and magnetite, the main diluents of the concentrate, could be expected to upgrade the selenium content byup to 80 per cent (relative) and .festwork carried out by Peko Mines N.L. on the basis of the recommendations of the mineralogical investigation confirmed that depression of chalcopyrite in the flotation circuit could produce significantly improved selenium grades.These investigations illustrate the application of mineralogy to the practical problems of improving plant performance. Mineralogy has an important role to play in providing quantitative data on which the metallurgist can evaluate plant performance and base a programme of metallurgical testwork, and early mineralogical investigation of feeds, concentrates, middlings, and tailings can lead to the avoidance of much abortive and expensive metallurgical testwork.