Multiple Discriminant Analysis in Type 2 Diabetes Melitus Outpatient in X Hospital Firman Pribadi Abdul Ghofir ABSTRACT This research aimed to analyze the relationship between the complication rates, the hospital profit rates to the cost of hospital treatment of patients with DM type 2 in outpatient of Hospital X. Statistically Test from Multiple Discriminant Analysis is done to predict the changes in the rate of complications in future. The results showed that the costs incurred by hospital is higher than the revenues received by the hospital especially for patients DM type 2 with more than one complication. More complications, will be higher for the cost to issued. For the average cost of antidiabetic medicine with the insuline, the results showed that all of the costs associated with many complications either uncomplications, the costs higher than the revenues received by hospital. For Further ,the results of test complication variable associated positively and significantly with non-insulin medicine costs variable, variable medicine with insulin, and many variations of medicine. The results of hospital profit test with the cost showed that the hospital profit rates correlated positively and significantly on the 1% to the total cost, but is not significant with non-insulin medicine costs and the medicine with insulin costs. This result explains that the profit levels did not correlate with the level of the medicine cost because the hospital profit obtained from the total cost issued by the payer of medical expenses. Then Multiple Discriminant Analysis test to predict the likelihood of occurrence in future indicates that there may be changes in the rate of complications from patients with DM type 2.Prediction results showed that there was a higher percentage of patients with DM type 2 will sustain the complications to a higher level, while the percentage of degression the complications rate of patient with DM type 2 is very low. This is understandable because the peregrination of patients with DM type 2 is likelihood very large to sustain the complications in future with predictive classification accuracy is 62.2%. Key Word: Treatment cost, Hospital Revenues, Multiple Discriminant Analysis