Spreading of a drop of an emulsion made with milk proteins on air/water interfaces was studied. From an unheated emulsion, all oil molecules could spread onto the air/water interface, indicating that the protein layers around the oil globules in the emulsion droplet were not coherent enough to withstand the forces involved in spreading. Heat treatment (90°C) of emulsions made with whey protein concentrate (WPC) or skim milk powder reduced the spreadability, probably because polymerisation of whey protein at the oil/water interface increased the coherence of the protein layer. Heat treatment of emulsions made with WPC and monoglycerides did not reduce spreadability, presumably because the presence of the monoglycerides at the oil/water interface prevented a substantial increase of coherence of the protein layer. Heat treatment of caseinate-stabilised emulsions had no effect on the spreadability. If proteins were already present at the air/water interface, oil did not spread if the surface tension (γ) was <60 mN/m. We introduced a new method to measure the rate at which oil molecules spread from the oil globules in the emulsion droplet by monitoring changes in γ at various positions in a 'trough'. The spreading rates observed for the various systems agree very well with the values predicted by the theory. Spreading from oil globules in a drop of emulsion was faster than spreading from a single oil drop, possibly due to the greater surface tension gradient between the oil globule and the air/water interface or to the increased oil surface area. Heat treatment of an emulsion made with WPC did not affect the spreading rate. The method was not suitable for measuring the spreading rate at interfaces where surface active material is already present, because changes in γ then were caused by compression of the interfacial layer rather than by the spreading oil. Copyright © 2002 Elsevier Science B.V.