<p>Previous studies on ecosystem service supply–demand (ESSD) balance have focused on sufficient socio-ecological factors, while overlooking necessary conditions. This gap limits policy insights, restricting them to correlations rather than identifying the essential prerequisites for achieving balance. To fill this gap, we integrated Geodetector and necessary condition analysis (NCA) to examine driving factors of ESSD balance from both sufficient and necessary perspectives. Using Jiangxi Province in southern China as a case study, we first quantified the spatiotemporal patterns of ESSD balance and then systematically diagnosed its drivers, with a particular focus on revealing heterogeneity between ecological and non-ecological counties. The results indicate that: (1) Between 2000 and 2019, the average ES supply across Jiangxi Province decreased by 12.96%, with high-supply areas located in peripheral mountainous regions, while the average demand surged by 56.25%, with high-demand areas concentrated in municipal districts. (2) The proportion of counties in ESSD deficit expanded from 24.2% to 45.1%, and the deterioration rate was significantly faster in non-ecological counties than in ecological counties. Spatially, surplus hotspots and deficit coldspots remained relatively stable. (3) Forest ratio, DEM, and slope were identified as key factors, exhibiting both high explanatory power (Geodetector q &gt; 0.48) and necessity (NCA d &gt; 0.17), with significant heterogeneity in their driving intensity across different county types. This study therefore offers&#xa0;provides a reference for implementing zonal and categorized precision ESSD management in Jiangxi Province and other similar regions.</p>