Abstract: Orderly restoration of cultivated land is essential to regulate cultivated land use for national food security. However, it is lacking in natural land endowments, particularly for the long-term stable use of restored cultivated land. It is often required to more rationally determine restoration priorities for the high stability of post-restoration agricultural use. In this study, a sequential zoning was developed for cultivated land restoration using land quality and site conditions. A study area was taken as Puyang City, Henan Province, China. An evaluation framework was constructed using the Land Evaluation and Site Assessment (LESA). The framework also consisted of Land Evaluation (LE) and Site Assessment (SA) subsystems. Among them, the Land Evaluation (LE) subsystem was used to characterize the natural quality of restorable land parcels, including obstacle-layer depth, soil profile configuration, top-soil texture, soil organic matter content, soil pH, salinization degree, irrigation guarantee rate, and irrigation water source. The Site Assessment (SA) subsystem was designed to evaluate the external conditions affecting the long-term stable utilization of restored cultivated land, including location advantage, economic benefit, cultivated land use condition, landscape value, and environmental index. Multiple data sources were integrated, including land use, restorable land category, agricultural land grading, and socio-economic data in 2022. The LE subsystem was scored, according to agricultural land quality grading, while the SA subsystem was calculated after indicator standardization and the entropy weight. A coupling coordination degree model was also introduced to identify the relative contribution of land quality and site conditions. The interaction between the LE and SA subsystems was quantified to determine the optimal dynamic weight ratio. According to the optimal weights, the integrated LESA score was calculated for each restorable land parcel. The natural breaks method was then used to classify the restorable cultivated land into different restoration priority zones. The results showed that the optimal weight ratio between LE and SA was 6:4, and the coupling coordination degree reached 0.913, indicating a high level of coordination between land quality and site conditions. The land quality provided the fundamental basis for cultivated land restoration, whereas the site conditions were important constraints to maintain stable use after restoration. The integrated LESA evaluation scores of restorable cultivated lands ranged from 52.28 to 91.61. The restorable cultivated land was divided into four categories after zoning: priority, secondary priority, general, and lagging restoration zone. Their areas were 25.63, 91.85, 79.06, and 113.68 km², respectively, accounting for 8.26%, 29.61%, 25.48%, and 36.65% of the total restorable cultivated land area, respectively. Spatially, the priority and secondary-priority restoration zones were distributed around urban built-up areas and county hinterlands, where there were relatively favorable land quality, irrigation conditions, accessibility, and cultivated land connectivity. In contrast, the general and lagging restoration zones were located along the Yellow River and the old Yellow River course, where the restoration priority was reduced by soil salinization, land fragmentation, weaker production conditions, and ecological constraints. The natural land suitability was integrated with the possibility of long-term stable utilization, thereby limiting the restoration decisions on land quality or suitability evaluation. The finding can provide a quantitative basis to prioritize differentiated restoration of farmland to support its sustainable resources.