The patient's medical record data on inpatient mental obtained from database of hospitals which is devoted namely mental hospitals, with the existence of medical record data encourages researcher try to conduct research using these data. The method used to Sequential Patterns mining of the return of patient formed by mental patient diagnosis algorithm Sequential Pattern Discovery using Equivalent classes (SPADE). This algorithm is used for the data in the form of the time sequence of transaction, and the results obtained in the form of pattern rules. By using the probability of events on a few items (diagnosis) suffered or experienced one patient from the overall results of the diagnosis in the medical record or which is called support of 0.02 and the probability of event on a few items (diagnosis) suffered by one patient or called confidence of 0.6 obtained pattern rules arrival which are formed from the results of the diagnosis of the patient as much as 6 recommended rules. The existence of this rule can be considered the hospital in the future to give a diagnosis of the patient on the patient's arrival. In additional, it can improve the quality and availability of screening tools, medicines and medical staff. The existence of rule’s pattern formed, it can be used as the information from the hospital to the patient's family to be aware of patients diagnosed with the disease returned to the hospital 
Keywords: Medical Record Data, Sequential Patterns Mining, SPADE