Radio frequencies refer to the electromagnetic energy that we transmit the identification information from tags to its reader. Radio Frequency Identification (RFID) transmits the data without line of sight. RFID tags are small, wireless devices that help identify item automatically and indicating unique serial number for each item. However, counterfeiting in supply chain management likes cloned and fraud RFID tag bring the impact to the organization and social when attackers want to gain illegal benefits.Organizationsarelosing a lot of money and trust from userswhen counterfeiting occurred. Furthermore, RFID data nature characteristics faces the issues likes RFID just carry simple information, in-flood of data, inaccuracy data from RFID readers and difficulties to track spatial and place. We propose to use clustering algorithms in order to detect counterfeit in supply chain management. We will apply various clustering algorithms to analyzed and determine every attribute in the dataset structure pattern. Based on evaluation that have done, we found that Farthest First is the best algorithm for 1000 (small data) and 10000 (bigger data). However, the values of false negative in data still quite high and it is dangerous if RFID scanner misread the cloned or fraud tags become genuine tags. Hence, we applied cost algorithms to reduce false negative values.