Local farmers are key stakeholders in the "Grain for Green Program," with their perceptions and willingness to re-enroll playing an important role in implementation of the next stage of this program. To discern local farmers' heterogeneous responses to the program and discover the reasons behind their responses, sustainable livelihood framework was utilized in the present study to analyze the essential elements having greatest significance on the farmers' responses. The sustainable livelihood framework includes five types of capital: natural, human, physical, financial, and social, with each capital composed of various indicators. Logistic regressions were utilized to identify essential indicators encompassing close relationships with local responses. The results revealed that 77% of respondents were willing to re-enroll in the next stage of the Grain for Green Program. The types of capital they possessed were significantly different between those who were willing to re-enroll in the program and those who were not, implying that capitals have distinct effects on respondents' willingness to re-enroll. Specifically, natural capital had a negative effect on respondents' willingness to re-enroll in the program, while financial capital and social capital had positive effects. Among the indicators composing the different types of capital, farmland size had the largest negative effect (β=-23.041) and cash income had the largest positive effect (β=38.591). In addition, the number of village committee memberships in the household and labor education attainment had positive effects on respondents' willingness to re-enroll (β=13.625 and β=7.717, respectively). With regard to cash income, it was demonstrated that the ratio of farm income to total cash income had negative effects (β=-13.627) on respondents' willingness to re-enroll, while the ratio of payment and non-farm income to total cash income had positive effects (β=13.603 and β=11.996, respectively). Furthermore, spatial distribution of the number of respondents willing to re-enroll into the program was allocated unevenly, with villages located further from the main road having a greater number of respondents willing to re-enroll than those who were closer to the main road. The above results disclose the complicated background related to farmers' willingness to re-enroll in the Grain for Green Program, with the underlying mechanisms discussed. The conclusions from the present study are that local dependence on the agricultural industry needs to decrease, while education attainment, payment and non-farm income, and information accessibility need to improve to implement and sustain the program. In addition, it is necessary to take into account spatial differences in sustaining or advancing the program. Larger numbers of farmers' residing in remote villages are willing to re-enroll in the program and their farmland and forested land is suitable for Giant Panda habitat. Therefore, it is recommended to advance the next stage program in remote villages, while sustaining the program in villages close to the main road.