Abstract: Knowledge expression and disease diagnostic reasoning are essential for the development of decision support systems for diagnosis and management of plant diseases. In this study, plant disease diagnosis knowledge was collected and summarized. Numerical expression of the disease diagnosis knowledge was achieved by using the principle of fuzzy mathematics and the multiple proportion method of evaluation. The numerical expression of the disease diagnosis knowledge was then combined with the "object - attribute - value triples act" (referred as OAV act) and the production rules to achieve effective expression of the disease diagnosis knowledge, and based on which,, two inference models for one-step diagnosis and two-step detailed diagnosis were developed by using fuzzy reasoning method to simulate the diagnostic reasoning of experts. The best-first search method and C # computer language were used for developing and programming the inference models. The knowledge expression and disease diagnostic reasoning models developed in this study can be used for the development of decision support system for disease diagnosis and management of greenhouse fruit and vegetable crops.