A safe and reasonable application of pesticides expert system was designed with B/S structure which had six sub-systems such as pests querying, pest identifying, decisionmaking of control methods, pesticide recommendation, knowledge management and expert consulting. Ontology technology was applied to model expert knowledge in order to achieve high intelligent reasoning and the sharing, and the knowledge reusing. Furthermore, a multi-factor model of pesticide evaluation was designed for comprehensively considering various factors that affect the reasonable selection of pesticides. The implemented prototype system verified that the solution was low-cost, high scalable, user-friendly and efficient. It is widely prospective in various vegetable productions and will supply a guideline for other crops production.