In experiment 1, potential intake rates of several dried forages, varying widely in nutritive value, were measured by offering each alone to six hungry sheep for eight, separate 1-min periods. Intake rates ranged from 5.5 to 26.1 g/min, but, within each forage, it varied little between sheep (c.v. 13%) or between measurements with the same sheep (c.v. 8%). The forages were then offered in pairs in separate containers and preference for one forage was defined as the percentage of total intake derived from that forage. With the exception of a dried clover pasture, preference for a forage was strongly related to the rate at which it could be eaten. Reducing the length of wheaten straw particles from 30 to 10 mm increased intake rate from 5.5 to 12.4 g/min and resulted in an absolute preference for the short material. The mean preference for a forage over all comparisons was more strongly correlated with intake rate of the forage (r2 = 0.87) than with in vitro digestibility of organic matter (r2 = 0.30). In experiment 2, two wheaten straws and two wheaten hays were each chopped to two lengths and, within each forage, mixed (w/w) in the short : long ratios of 0 : 1, 1 : 2, 2 : 1 and 1 : 0. Mean intake rates were 7.3 and 12.5 g/min for the two straws and 15.7 and 23.2 g/min for the two hays. Within each forage, all mixtures were offered in pairs to establish preference. The slopes of regression equations relating preference (%) to intake rate (g/min) were 21.4, 17.0, 5.9 and 2.7 (%.min/g) for the four forages as mean intake rate increased from 7.3 to 23 2 g/min, which indicates that discrimination between forages with the same difference in intake rate was greater when mean intake rate was low. In experiment 3, finely ground clover pasture, lucerne hay, wheaten hay or wheaten straw were added to a base diet of either chopped lucerne or wheaten straw in the ratio 1 : 10. Although the additive had little effect on intake rate, sheep preferred lucerne to clover pasture and wheaten hay to wheaten straw. Equations developed from the results of experiment 2 to predict the preference for one forage over another on the basis of their relative intake rates, showed fair agreement with observations from experiment 1, particularly when the effects of acceptability factors such as taste, odour or feel were translated into differences in potential intake rate.