In the context of global climate change, the widespread advance of spring phenology in temperate vegetation of the Northern Hemisphere has been shown to be closely related to climate change in the previous year. Moreover, the previous year"s warming climate has been shown to extend the growing season of temperate vegetation and enhance net photosynthetic rates, thereby increasing net primary productivity (NPP) and indirectly leading to the advancement of spring phenology in the subsequent year. However, the relative intensity of the direct and indirect impacts of the previous year"s climate on the subsequent spring phenology remains unclear. This study utilized long-term datasets on the start of the growing season (SOS), NPP, and meteorological variables for temperate vegetation in the Northern Hemisphere. By employing partial correlation analysis and structural equation modeling, we investigated both the direct influence of the previous year’s growing season climate on SOS and its indirect effects mediated through changes in vegetation productivity. The findings revealed that: (1) The length of the growing season played a predominant role in influencing vegetation NPP, being the dominant factor of NPP changes in over 50% of the areas studied. (2) Over the period from 1983 to 2014, the SOS of temperate vegetation in the Northern Hemisphere advanced at an average rate of 1.6 days per decade. Pre-season temperature had the greatest explanatory power for SOS, accounting for an average of 52.5% of its variation. The explanatory power of the previous year’s NPP on SOS ranged from 17% to 50%, comparable to that of spring precipitation and radiation in the subsequent year. (3) The previous year"s growing season climate not only indirectly influenced the subsequent SOS through NPP changes, accounting for 10.5% of the SOS variation, but also directly impacted the SOS in the subsequent year, explaining 19.7% of its variation. The direct impact was twice that of the indirect effect. These results highlight the significance of the previous year"s climate change and vegetation productivity on subsequent spring phenology, providing valuable insights for future accurate predictions of spring phenological events.