Phenology is the key indicator of climate change,and the rapid development of remote sensing provides a new method to monitor phenology.It is really significant of acknowledging the differences in extracting vegetation phenology to assess the availability of remote sensing product in monitoring phenology. Taking three northeast provinces of China as study area,our study exploits asymmetric Gaussian function fitting method to smooth the data,and uses dynamic threshold method to extract the start of the growing season(SGS),the end of the growing season(EGS),and the length of the growing season(LGS)of MODIS, CYCLOPES and GLASS leaf area index data product.The results show that MODIS and GLASS data product have similar results in extracting SGS,EGS and LGS,and the consistency between MODIS and GLASS data product is better on the whole;The SGS extracted from CYCLOPES is later than MODIS and GLASS data product in the overwhelming majority of cases,but the EGS is earlier than MODIS and GLASS data product,thus it causes the LGS shorter than the other two data products.The validation of phonological phase by taking advantages of the phonological observation data show that the SGS from MODIS and GLASS data product is similar to the phonological observations,and the EGS is a litter later than the phonological observations;The SGS and EGS from CYCLOPES data product are similar to the phonological observations and it indicates that the SGS and EGS from CYCLOPES data product in woodland are much more reliable.