The research regards the diagnostic data of the patients’ disease as the mining data source. Each diagnostic data includes patients’ symptom and sickness. The paper regards a certain patient as the mining target, considering the symptom weighted situation, using the clustering method in the data mining to dig the tendency of patients’ disease. In addition, the paper combines a group center formed by the patients’ symptom and designs a symptom weighted clustering method to satisfy the diagnostic data of minimal symptom similarity which belongs to the clustering. Later, the disease item whose number is the maximum can be found out in the clustering and the tendency of patients’ sickness. The methods proposed in the paper design and build a diagnostic system of patients’ sickness. The mining results of system can offer some useful referent information for those people check the sickness tendency of patients’ disease or those medical staff whose clinical experience is not enough confirms the disease diagnosis.