Biometric recognition systems are widely used for automatic authentication in different applications. The main goal of each biometric system is to accurately recognize individuals in verification or identification tasks. Biometric template protection (BTP) schemes provide protection for biometric templates against different security and privacy attacks. In addition to the protection provided by these schemes, it is important that the protected system has satisfying recognition accuracy. However, compared to unprotected biometric systems, we can consider different scenarios for a protected biometric system, and therefore we need to evaluate the recognition performance of the system in these different scenarios. This chapter presents an overview of the evaluation of protected biometric systems. We focus on standardized evaluation of biometric systems and discuss different scenarios in which the performance of a protected system should be evaluated, for both verification and identification (open-set and closed-set) tasks. Moreover, we discuss other criteria, such as complexity and memory consumption, as well as recognition bias, that should be taken into account when evaluating protected systems.