Perennial ryegrass is the dominant sown grass species in New Zealand, however its persistence in pasture is an issue for hill country farms. Deferring the grazing of a pasture over the reproductive cycle of ryegrass may influence changes in the above and below ground biomass of pasture plants and offers a low input and sustainable method for farmers to manage dry matter yields and quality of feed supply. The aim of this study was to investigate the above and below ground changes in biomass and tiller production of perennial ryegrass under different lengths of deferment, after which standard rotational grazing or simulated grazing was resumed. Two trials were set up to investigate this. A field study was undertaken in an established ryegrass/clover-based pasture with three grazing treatments. Standard rotational grazing already in use on the field site (Def x0), a period of deferment over late spring/summer from November to February (Def x14) and a longer period of deferment from late spring to autumn (Def x27). After deferment, standard grazing was resumed. Above and belowground biomass, tiller densities, nutritive value, pasture composition and ground cover were measured. A glasshouse-based study investigated the effects on perennial ryegrass of different lengths of delayed defoliation and cutting treatments applied to simulate grazing. The control was cut when 2.5 to 3 new leaves per tiller had grown since the last cutting (Def x0); cutting of other treatments was delayed for an extra 4 (Def x4), 8 (Def x8) or 12 weeks (Def x12). After the delayed cutting, plants in each treatment were cut whenever there were 2.5 to 3 new leaves present per tiller. Both trials were affected by an extra-long period of rest between grazing or cutting treatments from the end of February to late April/early May due to Covid-19. This acted as an additional period of deferment for all glasshouse treatments and for Def x0 and Def x14 in the field. The main results of the field study were that deferring resulted in higher tiller densities in winter and more accumulated dry matter before the deferred was opened for grazing than the rotational grazed control. There was also less bare ground compared with the control during summer. Nutritive value was lower during deferred periods but returned to the same value as the grazed control after standard rotational grazing resumed. There were no differences between any of the deferred treatments in root biomass. The percentage of ryegrass in total dry matter or ryegrass ground cover did not differ significantly between treatments. In the glasshouse study, delayed defoliation increased leaf biomass, live tiller number and reproductive tiller number. Treatments with longer periods of delayed defoliation reach higher peaks in above ground dry matter and reproductive tiller number. There was a temporary increase in root biomass at depth for treatments with delayed defoliation in January. After standard defoliation resumed, Def x12 had a higher percentage of dead tillers and lower biomass by the end of the trial in July. There were no differences between treatments for live tiller number or root biomass once standard defoliation resumed. The results suggest that a period of deferred grazing on pasture may improve dry matter yield and tiller densities of ryegrass after normal grazing resumes. The longer deferred treatment accumulated more dry matter in autumn, however it is not certain how the longer grazing rest between February and May affected potential differences between the shorter and longer deferred period. Results from the glasshouse saw no lasting benefit to ryegrass once standard defoliation resumed. Deferment length had no significant effect on belowground biomass. It is concluded that while deferred grazing may offer some benefits to a pasture, the persistence of ryegrass will not be improved by deferring grazing if other variables such as heat, moisture stress or timing of grazing are limiting potential growth. However, due to the unforeseen consequences of the Covid-19 lockdown, further research is recommended.