A comparison of a disk meter, sward stick and capacitance meter was conducted to determine the prediction ability of different methods for estimation of herbage mass (HM) on timothy (Phleum pratense L.)-meadow fescue (Festuca pratensis Huds.) pastures. A total of 714 paired observations was made on each instrument and corresponding cut HM (<1 cm). The variation in HM was adequately explained by linear models. Comparison was made separately for each occasion (DAY) since DAY had a significant influence on the parameter estimates for the sward stick and capacitance meter. The r2 values varied between 0 and 0.91 and the residual standard deviations (RSD) between 132 and 1151 kg ha-1. The disk meter and the sward stick predicted most accurately HM kg ha-1, while the capacitance meter was the least accurate. The advantage of the disk meter was the stability of model parameters from day to day. The advantage of the sward stick is that the readings can be applied directly into practice. None of the methods is reliable enough when precise HM estimates are needed from small-scale experiments, but they are useful in large areas or when HM estimates tolerate larger errors.