Abstract: Designing and formulating an ecological compensation mechanism for farmland non-point source pollution can effectively promote the environmental management and protection of farmland. However, the efficacy of compensation policies depends on the levels of acceptance, response, and implementation of such policies by farmers. In order to study the influence of different ecological compensation standards on the optimization of farmers' production behavior, we examined the case of the Hani Rice Terraces in Yunnan Province, China. We divided the farmers into two groups who cultivate at high and low altitudes respectively, and built a multi-objective production decision model based on the multi-objective utility model. By increasing the compensation amount in the profit function and setting different compensation standards, we simulated the production behavior of farmers in two groups under different compensation standards and analyzed the impact of different compensation standards on farmers' planting decisions and welfare. The results showed that although the policy of ecological compensation was aimed at reducing the use of fertilizers and pesticides, the additional income thus generated had stimulated farmers' enthusiasm for agricultural production. Driven by the pursuit of economic benefits, farmers invested more in terms of labor, tended to adopt more complex but higher-yielding planting structures. With an increase in the ecological compensation standard, the sensitivity of farmers' planting decision to the standard gradually decreased. In this regard, however, the high-altitude group was found to be more sensitive to the standard than the low-altitude group, and their planting structure changed to a significantly greater extent than that of the the low-altitude group. Furthermore, the intensity of fertilizer and pesticide input of the high-altitude group decreased to a greater extent than that of the low-altitude group. When the ecological compensation rate reached 3 000 ¥·hm^-2, the area ratios of rice, maize, intercopped soybean, and intercropped maize were 60%, 4%, 18%, and 18%, respectively; and fertilizer and pesticide usage was reduced by 37%, 49%, 37%, and 44%, respectively. Ecological compensation standards ultimately impacted farmers' incomes by changing their cropping decisions and chemical inputs. In the high-altitude group, the total benefits of farmers initially decreased but subsequently increased with an increase in the compensation rate. When the ecological compensation rate was 1 650 ¥·hm^-2, the benefits reached an inflection point. When the rate exceeded 1 650 ¥·hm^-2, not only did the farmers reduce the use of chemical fertilizers and pesticides, they also got a relative higher income. However, for the low-altitude group, the total yields of rice, maize monoculture, and maize intercropped with soybean continued to decline, and the reduction in pesticide and fertilizer application had a more pronounced impact on the total yields. Moreover, the response of farmers in this group to ecological compensation was also less positive. In this study, we demonstrated that agro-ecological compensation policies aimed at limiting chemical inputs would incentivize farmers to change their cropping decisions to compensate for the losses caused by a reduction in chemical inputs. Despite such reductions, changes in cropping patterns gave rise to uncertainty regarding total chemical inputs and farm household welfare. The results of this study accordingly highlight the importance of paying attention to changes in farmers' behavior in different environment during the implementation of ecological compensation policies. Ecological compensation has a significant effect on farmers' production behavior, and this effect is related to the production environment.