Received February 11, 1977 Accepted for Publication September 7, 1977 One of the important findings of the Reactor Safety Study (RSS) was the identification of the risk due to an interfacing system loss-of-coolant accident (LOCA), i.e., failure of interfaces between the high-pressure primary system and the low-pressure injection system (LPIS). Because equivalent interfaces exist in all pressurized water reactors (although not necessarily with the LPIS), the U.S. Nuclear Regulatory Commission (NRC) has included in its Standard Review Plan three equally acceptable designs intended to decrease the risk due to potential interface failures by decreasing the probability of an interfacing system LOCA. The present analysis of the RSS system configuration is in general agreement with the RSS results; however, the RSS presented a linearized estimate of the exact result, such that the probability of occurrence is overestimated from 0 to 5 yr of plant life and underestimated for plant life beyond 5 yr. In addition, this analysis shows that the NRC design options are not probabilistically equivalent; probabilistically, these options vary by four orders of magnitude, and one option could be implemented in such a way as to yield a probability of occurrence greater than the RSS evaluated design. Finally, as a demonstration of the power inherent in the probabilistic methods, the analysis itself reveals the dominate system failure (gross check valve leaks) leading to a limiting design that eliminates this failure mode and reduces the probability to an insignificant level