Moisture and treading treatments were imposed on intact turves that were relocated to a glasshouse after being removed from three hill pastures of different soil fertility in the North Island of New Zealand. The experiment consisted of a 2-month stress phase, where the treatments were wetting (W), wetting and treading (WT), drying (D) and control (C). In this phase, herbage accumulation rate, tiller density and leaf extension rate were lower on the D turves, and herbage accumulation rate and tiller density were lower on the WT turves than for the C turves. Herbage accumulation rate was higher on the W treatment than on the C treatment. In the 2-month recovery phase, herbage accumulation rate and leaf extension rate on the D turves were higher than those of the C treatment. Herbage accumulation rate and tiller density took longer to recover on the WT turves but by the end of the recovery period tiller density on these turves exceeded that of the C turves and the original tiller densities on the WT turves. Changes (increase or decrease) in leaf extension rate were associated with the W treatment and tiller density with the WT treatment. Moisture was limiting on the D and C turves, but on the W and WT turves, where moisture was adequate for plant growth, nutrients were limiting, notably phosphorus on the W and WT turves and sulphur on the W turves. The D treatment turves recovered very quickly once the stress was removed but the WT turves were slower to recover. Under the experimental conditions applied, the hill pasture turves were more resilient to the drying treatment than the wetting and treading treatment.