<p>Understanding the spatiotemporal evolution of rural structures is critical to achieving global sustainable development. However, due to the lack of long-term, micro-scale data at the national scale, these dynamic changes remain insufficiently observed. Here, we innovatively introduce a molecular perspective that conceptualizes villages as “atoms” and village structures characterized by specific inter-village spatial connections as “rural molecular structures”. Inspired by isomer concepts in physics and chemistry, we quantify these structures using rural molecular potential energy as a topology-sensitive metric of inter-village organization. Analyzing 495,314 villages across 2465 Chinese counties during 2000–2020, we show that underdeveloped northwestern counties are dominated by star-like aggregates, where a few low- or mid-stage villages drive the growth of numerous distant secondary settlements under large urban–rural income gaps. In contrast, developed southeastern counties form modular networked chains, with mid- and high-stage villages sustaining dense, localized interactions and high systemic cohesion. Nationally, hierarchical distances expanded rapidly in 2000–2010 (≈2 km yr⁻¹ on average; ≈6 km yr⁻¹ between Stage 3 and Stage 2), but contracted in 2010–2020 (≈1 km yr⁻¹; Stage 3-Stage 2 ≈ 1 km yr⁻¹), indicating a shift from centrifugal expansion to centripetal consolidation. Furthermore, causal forest inference indicates that urban–rural integration increases rural molecular potential energy, facilitating factor circulation, weakening core monopolies, strengthening long-range complementarities over time, and steering lagging regions toward more networked configurations. By translating rural spatial organization into a scalable and topology-sensitive metric, our approach provides a quantitative basis for monitoring rural structural reconfiguration and refining adaptive governance and intervention strategies in a timely manner.</p>