Vegetation phenology, the timing and the length of growing season, is sensitive indicator in response of terrestrial carbon, energy and water cycles to climate change. Understanding the changes of vegetation phenology as well as its response to climate change is of significance for predicting climate changes and global carbon cycle. Vegetation phenology are estimated mainly based on reflectance-calculated vegetation indexes (VIs), such as normalized difference vegetation index (NDVI) and enhanced vegetation index (EVI). However, large uncertainty exists in performance of VIs in land surface phenology due to small changes in seasonal variations of evergreen confers. The methodology of SIF to estimate land surface phenology has frequently been applied in deciduous forest. However, monitoring the phenology using SIF has not been sufficiently investigated in evergreen coniferous forest. In this study, we assessed the ability of reflectance-based vegetation indices (NDVI, EVI) and SIF datasets in monitoring the gross primary production (GPP)-based phenology in a subtropical evergreen coniferous forest. The vegetation indices were obtained from MODerate-resolution Imaging Spectroradiometer (MODIS) products and the SIF is retrieved from Global Ozone Monitoring Experiment-2 (GOME-2) from 2007 to 2011. Our results indicated that the GPP-based phenology, such as the start of growing season (SOS), the end of growing season (EOS), and the length of growing season (LOS) of subtropical evergreen conifers were day of year (DOY) 63, DOY 324 and DOY 272. The phenological metrics derived from the SIF were later than those derived from eddy covariance GPP, and the time-lag SOS_SIF and EOS_SIF were 19 days and 2 days, respectively. The time-lag SOS from NDVI and EVI was about 31 days related to SOS_GPP; meanwhile the time-lag EOS from NDVI and EVI was 10-17 days. The LOS derived from remotely sensed indexes were all shorter than the LOS_GPP, and GOME-2 SIF behaved better in capturing the growing stage of evergreen conifers in subtropical region of China. Spring temperature had the highest correlation with SOS, while water condition and solar radiation were determinants of the EOS at Qianyanzhou station. These results suggest that SIF, containing information of light-use efficiency, can accurately monitor the phenology of evergreen conifers and has important implications for biosphere models.