Information on grassland quality is required to assess the impact of agricultural policy in Europe. Field methods are time-consuming and local in scope, but may provide a basis from which to extrapolate using remotely sensed data. For thirty-eight fields in the Lorraine region of France floristic descriptions, agricultural observations, biomass, ground radiometric data and SPOT HRV data were collected during June, July and September 1986 and April 1987. Agronomical and phytosociological classifications of grassland management type and by inference grassland quality, yielded similar results but the phytosociological method was preferred as it could be quantified. Ground radiometric data and SPOT HRV data for all four seasons were used successfully to discriminate grassland quality but stratification by grassland quality did not increase the degree of correlation between remotely sensed data and vegetation amount. This research highlights the potential that exists for using SPOT HRV data, in conjunction with a phytosociological classification of grassland quality, to classify and then monitor, the quality of grasslands in Europe.