Phenology change as a result of climate change has dramatic influence on terrestrial ecosystem and hydrological cycle. Phenology detection models based on remote sensing data are important tools for regional phenological study. The study area of this paper is the Inner Mongolia grassland with focuse on the grassland phenology. We used the wavelet analysis method and data from remote sensing and meteorological stations to investigate the applicability of logistic model for phenology detection under different climate conditions. The results indicate that in areas with the drought index HAI larger than 3 the life circle of grass is disturbed, and the assumption of the logistic model will not be satisfied anymore, so under this climate condition it will not be reasonable to use logistic model for phenology detection; in areas where HAI is between 2 and 3, the vegetation growth may be disturbed by drought sometimes, so pre-test should be carried out when applying the logistic model.