A perennial ryegrass (Lolium perenne L.)‐dominated sward on a well‐drained soil (Experiment 1) and a creeping bent (Agrostis stolonifera L.)‐dominated sward on a poorly drained soil (Experiment 2) were subjected to four treading treatments: control (C, no damage), light damage (L), moderate damage (M) or severe damage (S) to quantify the effects on herbage dry‐matter (DM) production and tiller density. In Experiment 1, treading damage was imposed in spring. In Experiment 2, one‐third of the site was damaged in autumn, one‐third in spring and one‐third in both spring and autumn. Both sites were rotationally grazed after treading treatments. Pre‐grazing herbage mass was measured eight times in Experiment 1 and seven times in Experiment 2 on each plot, and tiller density was assessed four times in each experiment. In Experiment 1, pre‐grazing herbage mass was reduced by 30% in S plots at the first harvest after damage, but cumulative pre‐grazing herbage DM production was not different between treatments (12·7 t DM ha−1). In Experiment 2, annual cumulative pre‐grazing herbage mass was reduced by between 14 and 49%, depending on intensity of treading damage event and season when damage occurred. Tiller density was not affected by treatment in either experiment. A perennial ryegrass‐dominated sward on a well‐drained soil was resilient to heavy treading damage. A creeping bent‐dominated sward on poorly drained soil requires a more careful grazing management approach to avoid major losses in cumulative pre‐grazing herbage mass production during wet weather grazing events.