The effect of heat treatment of milk on whey proteins, sulphydryl (reactive and total), disulphide groups, half cystine, volatile sulphur compounds and sensory properties was studied in order to relate changes in the flavour of milks to changes in chemical composition. The degree of heat denaturation of whey proteins increased generally in the following order: HTST treatment, direct UHT treatment, indirect UHT treatment, and the severest was laboratory heat treatment where milks were held at temperatures below 1000 C for several minutes. At the same heating temperature and holding time indirect UHT treatment was found to result in more whey protein denaturation compared with direct UHT treatment. The sensitivity of whey proteins to denaturation was confirmed to be In decreasing order: serum albumin, S-lactoglobulin B, S-lactoglobulin A and a-lactalbumin. Laboratory heat treatment resulted in a peak of reactive sulphydryl groups, followed by a decrease, as the heating time is prolonged. HTST treatment and UHT treatments (direct and indirect) increased reactive sulphydryl groups from nearly zero in raw milk to values dependent on the heating temperature. Generally, heat treatment caused a decrease in concentration of total sulphydryl groups, disulphide groups and half cystine. Hydrogen sulphide and dimethyl sulphide were identified in heat-treated milks. Dimethyl sulphide was found to be present in raw milk as well as in heat treated milks. The increase in concentration of each of the volatile sulphur compounds was found to parallel the intensity of heating to the extent that they appeared to be linearly related. Volatile sulphur compounds decreased rapidly during storage at 22°C compared with~slower decrease during storage at 4oC. The panel was unable to detect differences between direct UHT milk and pasteurised milk. On the other hand, most panelists could detect differences between indirect UHT milk and pasteurised milk. The panelists were not able to distinguish between milks processed, respectively, at 77oc, 830 C and 890 C for 20 seconds using HTST treatment. A significant number of correct judgments w~s observed Wl+~ 940 C milk. The statements describing the difference were: heated taste, strong flavour, sulphurous, burnt flavour and very pronounced cooked flavour. All these statements are related to the cooked flavour. A good linear correlation was clearly observed between reactive sulphydryl groups and the whey proteins very sensitive to heat treatment, eg S-lactoglobulin A, S-lactoglobulin B and serum albumin. There was also a good linear correlation between reactive sulphydryl groups and mean whey protein denaturation. It was found that hydrogen sulphide concentration increased with increasing concentration of reactive sulphydryl groups, and with decreasing concentration of total sulphydryl groups, disulphide groups and half cystine respectively. There was also a good correlation between hydrogen sulphide and S-lactoglobulin A, S-lactoglobulin B and serum albumin. The degree of cooked flavour, the degree of whey protein denaturation, the concentration of reactive sulphydryl groups, the concentration of half cystine and the concentration of hydrogen sulphide show good correlation and similar dependence upon the severity of heat treatment. The data also suggest that the cooked flavour threshold corresponds to: a) 59 per cent denaturation of S-lactoglobulin (A + B); b) reactive sulphydryl group concentration of 0.037 mmole per litre of milk; c) half-cystine concentration of 1.082 mmole per litre of milk; d) hydrogen sulphide concentration of 3.4~g per litre of milk.