As amajorproducer and exporter in the international market, China has a significant responsibility to ensure its food quality and safety andimprove worldwide nutrition and health of human beings. However, some scandals of food safety issues in the last decade have resulted ina negative influence on China’s trade performance and unveiled the inadvertence of food safety supervision. The residue of pesticides is considered as the most serious sourcing pollution, indicating the necessity to control risks in farming at the beginning of the supply chain. Theoretically, cooperativeshaveshaveadvantages in organizing farmers’production, enforcing internal supervision, and relieving regulatory pressures of the government.Previous studies have focused on the effects of standardization, process management and other direct control measures in qualitative or quantitative ways, but none of them have noticed the interaction effects between different control measures. Besides, the effects of indirect control measures related to human resource incentives and organizational culture construction remain unknown. To fill in this gap, an internal managementframework, consisting 11 management measures of outcome-, process- and social-control aspects, is developed in this paper. In this way, not only the single effect of each measure is estimated, but also the optimal combination of management measures can be identified through PCA analysis. Furthermore, by simply using the real value of farmer’s Pre-Harvest Interval (PHI) subjects to various crop species and pesticide brands, we use farmers’ PHI deviation distance to a standard PHI value to evaluate the performance of farmers behaviors regarding quality safety. By averaging the PHI value of pesticides for each disease of corps, farmers producing different cultivated crops can be compared. Based on a survey database of 100 cooperatives with 312 households, process-control is proved to be the most effective measure. In specific, pre-production management measure like unified input, on-production management measure like unified production standard, and post-production management measure like unified brand can significantly improve farmers’ PHI deviation. Technical training is the only effective measure in social-control, while outcome-control shows no significant effect on farmers’ behaviorsdue to immature organizational structure of cooperatives and limited management skills. With regard to combinationsof management measures, the optimal combination is made up of unified input, unified production standard, technical training, recording and collaboration. Therefore, cooperatives should adopt proper management measures according to their developing stages, and Chinese government need to attach more attention to providing cooperatives with financial supports and training opportunities for both managers and members.