Abstract: Abstract: In order to improve farmland quality, the basic condition of cultivated land and its social suitability must be taken into account simultaneously for high quality farmland evaluation and construction. Therefore, with these requirements and objectives, the evaluation index system and identification model of cultivated land comprehensive quality were established. The index system was from three impact factors including land fertility level, land leveling project, infrastructure in cropland. Thirteen specific indicators including surface soil texture, profile pattern, degree of salinity, soil organic matter, drainage conditions, et al. were selected and analyzed. The weight of each evaluation indicator was determined by Delphi method. With GIS spatial analysis tools, the comprehensive quality suitability for high quality farmland construction was discussed by using the attribute hierarchical model (AHM) and exploratory spatial data analysis (ESDA). On the platform of study on external environment of high quality farmland construction, the index of rural households' comprehensive willingness consisting of the willingness of farmland quality improvement, adjustment of blocks, and scale of operation was constructed. This paper explored the correction model based on the index of rural households' comprehensive willingness, and the index of cultivated land comprehensive quality was modified in terms of rural households' comprehensive willingness. According to the grades of cultivated land comprehensive quality and the data of rural households' comprehensive willingness, the patterns of "comprehensive quality-comprehensive willingness" of high farmland construction could be obtained. The comprehensive suitability for time sequence and zone partition of high quality farmland construction and the concrete project of construction were discussed by using the diagnosis model and the combination of relative limited types. At last, on the platform of the investigation on rural households' comprehensive willingness, Quzhou county was used as a case study. Conclusions from the study can be drawn as: 1) The global spatial autocorrelation of cultivated land comprehensive quality was significant at 0.05 confidence level. The spatial clustering phenomenon about cultivated land comprehensive quality appeared on the whole. Aimed at cultivated land comprehensive quality, the suitable area of high quality farmland construction was 22472.84 hm2, which accounted for 47.87 % of the studied area. There were 1298 construction blocks that were suitable for high quality farmland construction. These blocks were in three towns, Anzhai, Disituan and Baizhai; 2) The study zones were divided into three patterns, high quality-low willingness (HQ - LW), low quality-high willingness (LQ - HW), and medium quality-medium willingness（MQ-MW）; 3) The recent construction of high quality farmland should be arranged in zone one and the main directions were land leveling project and land block adjustment. The area of comprehensive suitability zone one was 13714.22 hm2, accounted for about 29.22% of the studied area. The comprehensive suitability zone two was mainly located in Nanliyue and Houcun. Land fertility level and organic matter were main limiting factors. The long-term construction should focus on the comprehensive suitability zone three. Leading limitation types were infrastructure in cropland and land fertility level.