1. A group of six cattle, three of which had a non-bloating history (group A) and had been ruminally cannulated for the previous 2 years, and three with a history of being bloat-prone (group B) and which had been ruminally cannulated only 3 months before the study, were fed ad lib. on chopped lucerne (Medicago sotiva) hay, lucerne pellets, or a 100 g chopped hay and 900 g rolled barley grain/kg diet over three periods of 30 d each. Flow of rumen digesta, by reference to CoEDTA and chromium-mordanted fibres, and foam production from samples of rumen contents were measured.
2. Samples of rumen contents (50 ml) from group A produced foam heights of 150 and 60 mm, 2 and 4 h after feeding respectively, compared with 240 and 150 mm for group B (P < 0.05).
3. The fractional passage rate of the 1-2 mm particles mordanted with Cr did not differ (P > 0.05) between groups.
4. The fractional outflow rates (FOR) for CoEDTA 0-2 h and 2-7 h after feed was offered were 0.205 and 0.160/h for group A and 0.093 and 0.086/h for group B respectively (P < 0.05).
5. Rumen-fluid FOR 0-2 h and 2-7 h after provision of feed were significantly (P < 0.05) inversely correlated (r -0.74 and -0.85 respectively) with the amount of foam produced from rumen contents at these times.