In batch flotation tests conducted on ores from the Merensky reef, changes in froth stability invariably occur with variations in the reagent suite. The main reagents are collectors (primary and secondary), activators, depressants and frothers. Since the particles entering and leaving the froth in a batch flotation system are continuously changing, the stability of the froth can vary. Under these conditions the simplest measure of froth stability is the measure of water recovery at a fixed froth height. The batch flotation system developed at UCT allows for the separation of gangue which is entrained relative to gangue which is floated. It has been found that the presence of naturally floatable gangue (NFG) leads to froth stabilisation, whereas the presence of hydrophobic sulfide minerals may lead to destabilisation of the froth depending on the hydrophobicity (contact angle) of the sulfide minerals. This can vary with ore type since particle shape and amount of particles present can influence the extent of destabilisation. At low depressant dosages sodium isobutyl xanthate (SIBX) always results in lower froth stability than sodium ethyl xanthate (SEX). The frothing nature of dithiophosphate leads to increased froth stability and the addition of copper sulfate results in destabilised froths. Increasing depressant dosage reduces the stabilising influence of NFG and the depressant type (guar gum or CMC) also affects froth stability. Frother can be used in an attempt to overcome the destabilising effects of high depressant dosage. This work examines the effect of variations in the reagent suite and uses water recovered at a fixed froth height as an indication of froth stability in order to analyse these effects on the recovery of sulfide minerals, floatable gangue and entrained gangue. (C) 2011 Elsevier Ltd. All rights reserved.