Leak detection systems (LDSs) play a major role in enhancing reliability and operability of oil and gas pipelines. They have the functional capabilities to detect, locate, and quantify leaks before they can cause devastating effects to the environment and operation. The performance of LDSs is typically influenced by three different types of failures that may have severe consequences: delayed detection, missed detection, and false detection of a leak. The consequences of these failures lead to extensive financial losses. For example, missed detection leads to oil spills and exposes operating companies to financial risk and destroyed image while false detection results in unnecessary deployment of personnel and equipment. To maintain safety of personnel and the environment and ensure operation continuity, LDSs should be assessed regularly. To fulfill this need, a probabilistic performance assessment scheme based on limit state approach for fiber optic LDS has been developed. The inherent uncertainties associated with leak detection and reporting capabilities are modeled to determine the LDS detection failure probability that combines two failure events: missed detection and delayed detection. Moreover, the probability of false detection is derived in terms of the lowest detectable change, the threshold. These three parameters establish the basis for an overall assessment scheme that can be used to predict the performance of the LDS. The proposed assessment scheme has been applied to a case study to demonstrate its usefulness and feasibility.