AbstractIn two separate feeding experiments using a total of twenty‐four individually housed Ayrshire cows six silages made from perennial ryegrass were offeredad libitumwith supplements of concentrates. In Experiment I herbage with a dry matter (DM) concentration of 225 g kg−1received either formic acid (‘Add‐F’) at the rate of 2·0 litres t−1or undiluted cane molasses at rates of 10, 20 and 30 litres t−l; the mean daily silage intakes were 9·54, 908, 9·27 and 9·49 kg DM per cow and the daily milk yields, corrected to 40 g fat kg−1, were 23·2, 22·3, 22·8 and 22·9 kg per cow respectively but none of the differences between the four treatments was significant. In Experiment 2 herbage with a DM concentration of 269 g kg−1received formic acid at a uniform rate of 2·6 litres t−1either with or without an additional application of molasses at 20 litres t−1; the mean daily silage DM intakes were 8·70 and 9·28 kg per cow and the daily fat‐corrected milk yields were 22·2 and 21·9 kg per cow respectively and were not significantly different. In both experiments the effects of the treatments on milk composition were small and not significant. It is concluded that there were no advantages in applying molasses to herbage treated with formic acid, and that the rate of application of molasses to untreated herbage which equated with the formic acid application was 20·30 litres t−lwhen assessed on the basis of silage