The process of making a dough, allowing time for maturation, dispersing the dough in water and wet sieving/washing to obtain a protein fraction and starch milk was studied using response surface methodology by changing the water to flour ratio in dough making (400-710 g kg (1)), maturation time (45-660s) and the type of flour. Two grades of bread wheat flour and durum clear flour were studied. The effects of aging at ambient temperature for up to 29 days and the addition of ascorbic acid at 100 or 500 mg kg(-1) flour on separation behaviour were also studied for freshly milled high-grade (65% extraction) bread wheat flours at constant maturation time, 600 s, and at optimum farinograph water absorption value. The quantities and dry matter contents of the protein fraction and starch milk were measured; a sample of starch milk was centrifuged to obtain decantate, tailings and prime starch fractions, and the dry matter contents of each were determined. All the dried samples were also analysed for protein content. The fractional recoveries of dry matter and protein in the protein fraction, prime starch, tailings and decantate were calculated for each experiment. The acid values of flour oils were also determined on some aged flour samples. The results indicated superior separation characteristics in high-grade wheat flour compared with lower-grade flours. The water to flour ratio was more influential than maturation time within the range studied. Contrary to the initial expectation, no statistically significant effect of flour aging was observed in the studies with no additive, and ascorbic acid addition was not found to improve the wet separation behaviour, the separation behaviour becoming even worse at the 100 mg kg (1)level. Acid value showed a slight increase with time. (C) 2002 Society of Chemical Industry.