Abstract: Based on the data of the flowering phenology for Prunus davidiana and daily average temperature (T_m), maximum temperature (T_max), minimum temperature (T_min), relative humidity (RH), precipitation (Pr) and sunshine hours (S) in Nyingchi meteorological station from 2001 to 2021. with linear regression method, pearson correlation coefficient and stepwise regression method, the change trend of flowering phenophase, its growing duration of Prunus davidiana, and the leading climatic factors were revealed, a flowering period prediction model was established. The results showed that (1) From 2001 to 2021, except for a delayed trend at the terminal flowering date (2.68 d/10a), the flowering period of Prunus davidiana were advanced by a rate of (0.65~7.27)d/10a in southeastern Tibet, in which the flower bud enlargement date advanced the most (p<0.001) and the flowering initiation date advanced the least (p>0.10). The length of flowering period was significantly prolonged, with an extension rate of 3.32d/10a. (2) During the flowering growing season, T_m and T_max increased weakly, while T_min did not chang. DTR did not increase much, ≥0 ℃ accumulated temperature increased significantly, S did not increase significantly,while RH and Pr decreased. (3) Before flowering, the main factor affecting the flowering phenology was RH or Pr, while the influence of T_m dominated after flowering. If each increasing (decreasing) of T_m by 1℃ in February, the initial flowering date and common flowering date would be advanced (delayed) by 3.45d and 4.27d, respectively. (4) The prediction accuracy of Prunus davidiana flowering period prediction model was 96.6% ~ 97.1%.