Four pigs fitted with a gastric cannula were fed on a wheat-flour-based diet (WF) and three oat-based diets, consisting mainly of oat flour (OF), rolled oats (RO) or oat bran (OB), for 1 week each. The stomach contents were collected quantitatively daily at 0·5, 1, 2, 3 or 5 h after feeding. The viscosity (mPa. s) of the liquid fraction of stomach contents 1 h after feeding was 1·7 with diet WF, 15 with diet OF, 30 with diet RO and approximately 400 with diet OB. The viscosity and the concentration of β-glucan in the liquid phase was to some extent determined by the dietary level of β-glucan in the diet. However, there was a trend towards a lower viscosity after longer exposure to the gastric juices. The correlation between logarithmic values for viscosity and concentration of β-glucan in the liquid phase of digesta was r 0·45. On centrifugation of digesta there was a higher proportion present in the sediment phase when the pigs were fed on diets with a higher content of soluble dietary fibre (DF), suggesting that the digesta was more coherent. This possibility was supported by the higher water-holding capacity (WHC) of the sediment. Feeding diets with oats containing a higher soluble DF content led to lower recoveries of digesta, PEG 4000 (liquid-phase marker), and the DF components β-glucan and arabinoxylan in the first hour after feeding. No effect related to the DF content of the diet was seen in the gastric emptying of starch and Cr2O3 (solid-phase marker). In conclusion, soluble DF from oatsincreased the viscosity of stomach contents and increased the ability of the dry matter to retain water. Higher levels of soluble DF led to higher recoveries of digesta, the liquid phase and DF itself in the initial stage of gastric emptying, whereas no effect was seen on the gastric emptying of starch.