In pasture-based systems farmers need to make daily management decisions to ensure that their livestock have good quality feed in adequate quantities. 2018 was a challenging year in Ireland due to a late start to grass growth in the spring and a drought in the summer. This situation highlighted the need to be able to predict grass growth weekly to help farmers manage their grass and better anticipate grass shortages in a changing climate. The grass growth modelling program at Moorepark, based on the MoSt Grass Growth Model, increased from 39 farms in 2019 to 78 in 2022 allowing a good representation of grass growth across Ireland. To predict grass growth, data from the farms participating in the project are extracted from PastureBase Ireland (PBI – a grassland decision support tool). Data includes number of paddocks and their area, grazing and cutting dates, number of grazing animals and feed supplementation, and nitrogen fertiliser application date and rate. Other data required by the model include the soil type for each paddock, and historical and forecast weather data (provided by Met Éireann, the Irish Meteorological Service). This paper will present the evaluation of four years of grass growth prediction by the latest version of the MoSt model compared to PBI farmer data using historical weather data. The number of available data points possible for comparison was 191,021 at the paddock level and 9,695 when averaged at farm level. At farm level, the average RMSE was of 13.5 (39 farms), 17.5 (56 farms), 15.7 (78 farms) and 14.3 kg DM/ha (78 farms) for 2019, 2020, 2021, and 2022, respectively. At paddock level (n=121,021), the error was higher and was 24.2 (1,295 paddock), 27.3 (1,880 paddock), 26.0 (2,471 paddock) and 24.8 kg DM/ha (2,478 paddock) for 2019, 2020, 2021, and 2022, respectively. The lower accuracy at paddock level is expected due to the difficulty in accurately replicating inter-paddock variability and the low reliability of the data entered into PBI at paddock level. While the predictions are not perfect, the accuracy at farm level is sufficient to make the MoSt model a useful tool for grassland management on Irish farms.