Quantification of biophysical parameters is needed by terrestrial process modeling and other applications. A study testing the role of multispectral data for monitoring biophysical parameters was conducted over a network of grassland field sites in the Great Plains of North America. Grassland biophysical parameters [leaf area index (LAI), fraction of absorbed photosynthetically active radiation (fPAR), and biomass] and their relationships with ground radiometer normalized difference vegetation index (NDVI) were established in this study (r²=.66–.85) from data collected across the central and northern Great Plains in 1995. These spectral/biophysical relationships were compared to 1996 field data from the Tallgrass Prairie Preserve in northeastern Oklahoma and showed no consistent biases, with most regression estimates falling within the respective 95% confidence intervals. Biophysical parameters were estimated for 21 “ground pixels” (grids) at the Tallgrass Prairie Preserve in 1996, representing three grazing/burning treatments. Each grid was 30×30 m in size and was systematically sampled with ground radiometer readings. The radiometric measurements were then converted to biophysical parameters and spatially interpolated using geostatistical kriging. Grid-based biophysical parameters were monitored through the growing season and regressed against Landsat Thematic Mapper (TM) NDVI (r²=.92–.94). These regression equations were used to estimate biophysical parameters for grassland TM pixels over the Tallgrass Prairie Preserve in 1996. This method maintained consistent regression development and prediction scales and attempted to minimize scaling problems associated with mixed land cover pixels. A method for scaling Landsat biophysical parameters to coarser resolution satellite data sets (1 km²) was also investigated.