The evaluation of cultivated land quality can provide support for the management of cultivated land quality. In order to objectively obtain the distribution of cultivated land quality in the province and reduce the evaluation error caused by time breakpoints, an evaluation method of cultivated land quality was constructed based on NPP inversion of longterm sequence remote sensing data. Firstly, rice and maize were identified by using timeseries LSWI and EVI data under the constraints of ploughing maps. Then, the MODIS MOD09A1 data from 2000 to 2010 were used, combined with the meteorological site data, the VPM model was used to calculate the NPP of rice and corn respectively, and the average NPP value was obtained for many years, reflecting the longterm growth of cultivated crops. The anomaly analysis method was used to eliminate the differences in crop types, and the evaluation results of the quality of cultivated land were obtained. Finally, with Jilin Province as the research area, the method proposed was empirically analyzed. The experimental results showed that the longterm sequence NPP cultivated land quality inversion results were consistent with the overall spatial distribution of cultivated land quality utilization. Among them, the inversion results were accurate in the central and northern parts of Jilin Province due to the concentrated arable land. The plaques in the southern mountainous areas were broken, and the cultivated land was mixed with the forest land, resulting in high inversion results. Therefore, the longterm NPP cultivated land quality evaluation method constructed was feasible for the evaluation of provincial and central contiguous arable land quality, and can provide scientific support for the evaluation of cultivated land quality in China.