By field trials, the canopy hyperspectral reflectance and chlorophyll content (SPAD) for winter wheat during 2010 and 2011 growth periods were measured by the ASD portable spectrometer and portable chlorophyll meter SPAD-502, respectively. The canopy spectral characteristics of different SPAD values were analyzed in different growth periods. The winter wheat SPAD estimation models based on normalized difference vegetation index (NDVI), ratio vegetation index (RVI) and wavelet energy coefficients were established in different growth periods. The results showed that green peak and red valley characteristics became more and more obvious with the increase of the SPAD. The SPAD estimation models based on NDVI performed better at the regreening stage, elongation stage, heading stage and filling stage with determination coefficients (R²) being 0.7957， 0.8096， 0.7557 and 0.5033, respectively. The winter wheat SPAD estimation models based on wavelet energy coefficients could greatly improve the SPAD estimation accuracy, with regression determination coefficients (R²) of the PVC estimation models based on high frequency energy coefficient and low frequency energy coefficient being 0.9168, 0.9154, 0.8802 and 0.9087 at the regreening stage, elongation stage, heading stage and filling stage, respectively.