Data Mining is a technique that can utilize large amount of data to dig up valuable information which is unknown before. This technique is a combination between statistic discipline merges with artificial intelligence and maachine learning based on database technology. PT CAR as large insurance company in Indonesia that serves a lot of health insurance policy holders. That can be realized because of processing of existing insurance policy holder with their applied claim. With this realization competency level can be raised. This resarch will be discussed about data mining technique with CRISP-DM (Cross Industry Standard Process for Data Mining) model as its development methodology which is adapted with this thesis research. Implementation done using Microsoft SQL. Server 2005 C# Net and for data cleansing process done by program written by author using C# programming language. This thesis will attempt to find feasible technique to filtering health insurance policy holder candidate which is most precise and efficient along with accuracy level of assessment that is done by the chosen data mining technique based on that analysis result. This research will ensure that data mining is helpful for insurance company to make efficiency of their taken total claim and also raise their competency level. Key Words: Data Mining