Abstract: 【Abstract】To clarify the impact of changes in meteorological factors such as temperature（TM），precipitation（TP）and solar radiation（SSRD）on the spatiotemporal variation of net primary productivity（NPP）in farmland under the background of climate change，which could provide reference basis for scientifically guiding agricultural production in Jiangsu.【Method】Based on MODIS remote sensing reflectance data，MCD12Q1 land use type data and ERA5 meteorological data，the improved CASA model was used to estimate the farmland NPP in Jiangsu from 2003 to 2023. The Theil-Sen median trend analysis（T-S trend analysis）and Mann-Kendall test（M-K test）were applied to explore the spatiotemporal patterns of farmland NPP in Jiangsu over the past two decades and their driving factors. Path analysis and Hurst index analysis were employed to reveal the contributions of meteorological factors to the variation of farmland NPP in Jiangsu at different scales.【Result】From 2003 to 2023，the interannual variations of meteorological factors and NPP in farmland across Jiangsu exhibited different trends. Specifically，TM displayed extremely significant increasing trend overall （P<0.01），with a linear trend rate of 0.04 °C/year，while the interannual changes in SSRD，TP，and NPP were not significant（P>0.05）. The spatial distribution of the annual average NPP of Jiangsu farmland showed a pattern of higher values in the north and lower values in the south. At a latitude of 34°N，the annual average NPP of Jiangsu farmland reached its peak（698.3 g·C/m²），decreasing symmetrically to the north and south；at longitude，the annual average NPP in the west of 121° E showed fluctuations（640.0 to 690.0 g·C/m²）. The interannual variation of NPP of Jiangsu farmland showed a significant overall declining trend（P<0.05，the same below），with notable spatial heterogeneity in the dominant factors influencing this variation. At the pixel scale，82.61% of the variation in Jiangsu farmland NPP were primarily influenced by TM，while 15.40% and 1.99% were mainly influenced by SSRD and TP respectively. As the analysis scale expanded，the dominant factors gradually shifted to SSRD. Regions in Jiangsu with significantly declining stable farmland NPP exhibited greater persistence，whereas regions with significantly increasing NPP were relatively weaker，leading to the conclusion that the future variation in farmland NPP acrossed most areas would likely continue the current trends.【Conclusion】Over the past two decades，farmland NPP in Jiangsu has generally shown a declining trend，but there exists considerable spatial heterogeneity among regions. In the future，farmland NPP in most areas of Jiangsu is expected to continue the current trend of variation. The driving factors behind the variation of farmland NPP also differ across various scales，with temperature and radiation being the primary influencing factors.