It is essential to have accurate data relating sward herbage mass (HM) to voluntary dry matter intake (VDMI) and in vivo organic matter digestibility (OMDIV) across the year. This will ensure that the production performance of animals grazing differing HM swards can be accurately predicted using in vitro and in situ methods. The objectives of this experiment were to (i) determine the OMDIV and VDMI of sheep offered perennial ryegrass differing in HM throughout the growing season and (ii) evaluate in vitro gas production (IVGP) and in situ indigestible neutral detergent fibre (iNDFIS) as methods of predicting OMDIV. A 3 × 3 Latin square design was used to investigate three HM treatments. The target HM (measured > 4 cm) of perennial ryegrass for the three HM treatments was 1200 kg dry matter (DM)/ha (Low), 2500 kg DM/ha (Medium) and 4000 kg DM/ha (High). The experiment was a Latin square design repeated in three time stages (TS) in 2013 (TS1, TS2 and TS3). A 3 × 2 incomplete Latin square design was used in early spring 2014 (TS4). In TS4 the HM were lower than in the previous TS at 1000 kg DM/ha (Low), 1800 kg DM/ha (Medium) and 2500 kg DM/ha (High). Each experimental period within each TS consisted of a 6 day adaptation phase and 6 day measurement phase; with 12 Texel wether sheep used. Sheep were fed ad libitum to ensure a 10% refusal rate. The interaction between TS and HM treatment had the greatest effect on chemical composition, VDMI and OMDIV. Low HM swards had the greatest VDMI and OMDIV. For each 1000 kg DM/ha increase in HM, OMDIV was reduced by 2.3% on average. This varied from 0.9% in TS4 to 2.7% in TS3. There was no difference between HM treatments in the rate of digestion of digestible NDF using the IVGP technique. In situ indigestible NDF concentration and OMD derived from IVGP (OMDGP) were highly correlated with OMDIV (r2 = 0.901 and 0.895 respectively) and moderately correlated with digestible organic matter intake (DOMI) (r2 = 0.633 and 0.652 respectively). It can be concluded that the interaction between HM and TS will greatly affect OMDIV and VDMI. Secondly OMDIV can be predicted by both iNDFIS and OMDGP with acceptable accuracy to allow for their use in future feed evaluations.