Models to predict neutral detergent fibre (NDF) and acid detergent fibre (ADF) of alfalfa before harvest have been developed in the USA, but have not been tested in northern Europe or above 50°N latitude. Our objective was to test predictive equations for alfalfa quality (PEAQ) and the New York prediction equation for alfalfa quality (NYPQ) for estimating alfalfa NDF and ADF in northern Poland. A total of 275 alfalfa samples, varying in height and maturity, were collected from twenty farm fields over 3 years. Observed NDF and ADF values were regressed on predicted values. The accuracy of PEAQ for predicting NDF and ADF in Poland was equal to that observed in the USA (r2 > 0·89; RMSE < 21·6). In most cases regression equations were not biased (slope = 1 and y-intercept = 0), and when there was bias it was minimal. The relationship between observed and NYPQ prediction of NDF was also strong (r2 > 0·88; RMSE < 23·0), but these equations were extremely biased. The results demonstrate that the PEAQ model provides a simple, rapid and reasonably accurate means to estimate fibre concentrations of alfalfa before harvest, appropriate for the growing conditions of northern Europe.