In 76 samples of Italian ryegrass, perennial ryegrass, timothy and red clover, harvested in different morphological stages, the crude fibre (CF) content was determined according to the method of Naumann and Bassler (1976) and neutral detergent fibre (NDF), acid detergent fibre (ADF) and acid detergent lignin (ADL) were determined according to the methods of Goering and Van Soest (1970). CF, NDF, ADF and ADL contents of the same samples were also determined with the use of filter bags and labelled as CFFB, NDFFB, ADFFB and ADLFB. In contrast to grasses, where the differences between NDF and NDFFB, ADF in ADFFB and ADL in ADLFB were small (in average the largest difference was 20 g kg-1 DM between ADF and ADFFB in perennial ryegrass), the differences between NDF and NDFFB, ADF and ADFFB, ADL and ADLFB contents in red clover were large (in average 46, 66 and 35 g kg-1 DM, respectively). The differences between CF and CFFB contents in grasses and red clover were not so large (in average teh largest difference was 17 g kg-1 DM in red clover). If the samples of red clover were rinsed with acetone before NDFFB, ADFFB and ADLFB contents determination (acNDFFB, acADFFB and acADLFB) the differences between them and NDF, ADF and ADL contents greatly diminished (in average to 15, 17 and 5 g kg-1 DM, respectively). The accuracy of the relationship between CF, NDF, ADF and ADL contents and fibre contents determined with the filter bags was measured with the coefficient of determination (R2) in standard error of the estimate (SEE). The best predictions for CF (R2 = 0.89, SEE = 1.73), NDF (R2 = 0.98, SEE = 1.73), ADF (R2 = 0.90, SEE = 1.73) and ADL (R2 = 0.69, SEE = 1.73) contents were those which contained as a single dependent variable CFFB, NDFFB, ADFFB and ADLFB of grasses and CFFB, acNDFFB, acADFFB and acADLFB of red clover. Results showed that the filter bag method is convenient for the determination of the CF, NDF and ADF contents, but is less suitable for the determination of ADF content.