Abstract: Biometrics recognition systems are widely used in many areas of application, so, security and privacy vulnerabilities have attracted the attention of the biometric community recently. The stored reference data are prevented from detecting special biometric information and the security of biometrics systems are enhanced against attacks such as identity theft and cross matching by using template protection techniques. This study focuses on a template protection algorithm that combines between the methods from Chaotic Shift Keying (CSK) and Visual Cryptography (VC). The key component of the algorithm is to use CSK modulation for coding biometric templates into chaotic signal. Then, using (2, 2) VC to generate two shares, these two shares are stored in two separate servers of the database, so that, the coded template can be detected only when both shares are available at the same time also, the identity of the private template does not detect through single share. It is noted that the generated shares should be robust, uniformly distributed, collision-free and statistically independent for improving authentication performance and avoiding information seepage. Biometric template protection techniques are capable of solving the problem of the vulnerability that undermines the biometric template.