The performance of the Renison tin concentrator was assessed in detail for two arrangements of the three available ball mills. With two mills used for primary grinding and the third for regrinding table tailing, the concentrator recovered 65.3 per cent of tin in the three tin concentrates produced by tabling, vanning and flotation at a calculated overall grade of 37.5 per cent tin prior to batch dressing. In the second arrangement one mill was used for primary grinding and two mills for separate regrinding of table tailing and sulphide concentrate. The metallurgical effective- ness of the change was somewhat obscured by unintended changes,i.e. an increase in the assay of the fresh feed from 1.38 per cent to 1.1+6 per cent tin and unresolved changes in the sulphide rougher flotation and tabling. However, using the results from a detailed size by size analysis of the behaviour of ore components in appropriate parts of the plant, it proved possible to show that, had the two unintended changes not occurred, up to 5 per cent