AbstractTwo experiments were conducted to determine the effect of pre‐harvest treatment by formic acid spray on the moisture concentration (MC) of perennial ryegrass (Lolium perenne) in vegetative growth. In Experiment 1 aqueous solutions containing 0.080, 0.148 and 0.259 kg formic acid per litre were applied to a perennial ryegrass sward (fresh matter yield 21.0 t ha−1) at a constant volume of 41.2 litres t−1to give active ingredient application rates of 3.5, 6.0 and 11.5 kg t−1 respectively. In Experiment 2 similar solutions were applied to a perennial ryegrass sward (fresh matter yield of 6.9 t ha−1) at 35.2, 19.0 and 10.9 litres t−1 respectively to give a constant formic acid application of 2.8 kg t−1. The changes in MC and chemical composition after treatment with acid were compared with those in untreated plotsIn both experiments treatment with acid significantly reduced the MC (P < 0.001) within 5 h of treatment from 4.4 to 3.0 kg per kg of dry matter (DM) in Experiment 1 and from 3.0 to 2.6 kg per kg DM in Experiment 2. There was no effect of level of acid applied in Experiment 1 or of the volume of water applied with the low level of acid in Experiment 2 on the changes in MC of acid‐treated plots. There was however a significant (P < 0.00l) difference between the MC of plots treated with acid and those untreated which was maintained in Experiment I for up to 9 d even during rainfall.There were no differences between the chemical compositions of samples from acid‐treated plots in either Experiment I or 2 or between untreated and acid‐treated plots in Experiment 2. However there were significant differences between untreated and acid‐treated plots in Experiment 1 for the content of nitrogen (N), water‐soluble carbohydrate (WSC) and digestible organic matter (DOM). There were changes in the content of N, WSC and DOM with time in both experiments. In Experiment I there was an interaction between sampling time and treatment for both N and WSC but not for DOM. Compared with untreated plots, treatment with acid at all levels accelerated the loss of WSC and delayed the fall in N content. There were no such interactions in Experiment 2.