Hepatitis is an inflammation of the liver. There are five types of hepatitis virus. Meanwhile, hepatitis B and hepatitis C virus can progress into liver cancer. Someone that is suspected to have hepatitis, can do a laboratory research to gain information about their condition. The data and information from the laboratory could be used to build a program which will predict the outcome for new data with similar symptoms of hepatitis B and hepatitis C. In this paper, the outcome of this problem was predicted using K-Means Clustering method based on the early information or data from hospital's laboratory. K-Means Clustering was a clustering method which provide 84.85 % accurate prediction for new data with similar symptoms with hepatitis B and hepatitis C. This method will decide whether a new data with similar symptoms to hepatitis B and hepatitis C will be classified as hepatitis B or C. Thus, new data and information about their health condition considering to hepatitis B and hepatitis C could be effectively presumed.