The emulsifying abilities of 12 soy protein isolate samples were studied in relation to their solubilities (Sol) in different solvents and their sulfhydryl contents (SHL). Samples were obtained at pilot plant under different heat and reducing treatments, according to a central composite design 22 + star with a triplicate central point, resulting in 11 samples. An untreated sample (standard) was also included. The independent variables were heat treatment temperature (T) and sulfite concentration (C
Na2SO3) while the responses were Sol in water (Solw), Sol in phosphate buffer (Sol buf), Sol in 2-mercaptoethanol (Sol2-Mer), Sol in SDS (Sol SDS) and SHI. The results indicated a good fit to a second order polynomial equation for Solw, Solbuf' Sol 2-Mer and SHL, while in the case of SolSDS the only significant term was the linear one for T. The dependence of emulsion ability (A
500 as a measure of interfacial area) with time (t) and protein dispersion concentration (C) was evaluated using response surface methodology (RSM). The A
500 showed a very good fit to a second order polynomial equation with R
2 higher than 89%. The structural differences among samples caused by different treatments were well explained by the A
500. Simultaneous heat and reducing treatment under mild conditions (T = 60-65 °C and C
Na2SO3 = 0.2- 0.4%) were able to open the protein structure, increasing SHL. Samples highly soluble in every sol vent (without aggregation effects) showed the highest emulsion ability.