The present research aims at utilizing hyperspectra information to estimate the content of endogenous hormones for predicting the bamboo flowering. The authors selected the giant panda’s staple-food bamboos Bashania fargesii, Fargesia qinlingensis and Fargesia dracocephala in different flowering situations to measure their hyperspectral reflectance in Foping Nature Reserve. The enzyme-linked immunosorbent assay was used to measure the contents of four endogenous hormones: Gibberellin (GA3), Auxin (IAA), Zeatin nucleotide (Zr) and Abscisic acid (ABA), and non-parametric test was applied to analyze their differences. Then, we built the responding relationship models between hormones and hyperspectral reflectance. The results showed that (1) GA3 contents of B. fargesii and F. dracocephala were higher in the flowering bamboo than in the non-flowering bamboo, while F. qinlingensis showed the opposite result; IAA and ABA contents of three bamboos were all lower in non-flowering bamboo than in the flowering one; Zr contents for three bamboo species were higher in the non-flowering bamboo than in the flowering one. (2) Except that B. fargesii’s IAA and Zr had significant relationships with original spectral reflectance among some bands, other hormones of bamboos all showed the poor relationships with the original spectra; however, some specific bands of all four hormones of three bamboos had the very significant relationships with the first derivative reflectance. (3) The precisions of linear models were higher than the Parabola models. The authors concluded that the close relationships between the hormones’ contents and spectra reflectance can be used to forecast the flowering of panda bamboos.