ABSTRACTThe increased risk of surprise attacks resulting from changed Navy missions (Hansen 2000; Hansen and Wagner 2001) makes it desirable to consider the possible development of improved passive protection systems. It is not obvious, however, which direction such development should take for optimum returns. This paper provides a tool to compare the effectiveness of different general approaches in terms of the reduction in out‐of‐action probabilities they can provide. A method is described to estimate the probabilities for different ships attacked by weapon salvos, and an illustration of the method is given by applying it to two general conceptual ship types, each with three different postulated levels of passive protection, all exposed to salvos varying from one to four weapons in size. Conclusions are given on the basis of the illustrated example, with the caveat that they should be revisited as additional results are developed. The feasibility of the different approaches to passive protection has to be determined by research and development.