Detection of grassland canopy chlorophyll fluorescence (Chl-F) conducted with an imaging spectroradiometer provided evidence of potential remote sensing estimation of steady-state Chl-F (Chl-Fs). Daily near-nadir views of extremely high spatial resolution hyperspectral images were acquired from a distance of 4 m for temperate montane grassland in the Czech Republic. Simultaneously, measurements of Chl-F and total chlorophyll content (Chla + b) were made on a single leaf at ground level were collected. A specifically designed ‘shade removal’ experiment revealed the influence of dynamic physiological plant processes on hyperspectral reflectance of three wavelengths: 532, 686 and 740 nm. Based on this information, the vegetation indexes R686/R630, R740/R800 and PRI calculated as (R532–R570)/(R532+R570) were tested for statistical significance with directly measured Chl-F parameters (maximum fluorescence yield, Fv/Fm; steady-state chlorophyll fluorescence, Chl-Fs and actual quantum yield, ФII). The grassland species under investigation were: Festuca rubra agg. (L.), Hieracium sp., Plantago sp., Nardus stricta (L.) and Jacea pseudophrygia (C.A. Meyer). The coefficients of determination (R2) for best-fit relationships between PRI-ФII and PRI-Chl-Fs, measured in the daily course, show a high variability of 0.23–0.78 and 0.20–0.65, respectively. Similarly, R2 for the R686/R630-ФII and R686/R630-Chl-Fs relationships varied between 0.20–0.73 and 0.41–0.70, respectively. The highest average R2 values were found between PRI and Chla + b (0.63) and R686/R630 and Chla + b (0.72). The ratio R740/R800 did not yield a statistically significant relation with Chl-F parameters.