This document is the final report of the Local Warning System Definition project performed for the Stanford Research Institute under Subcontract SRI 12675 (6300A-680). The report describes a method for determining an optimum mixture of public warning systems. Two types of input data to the optimum mixture method are developed: warning effectiveness standards based on empirically determined news dissemination rates and broadcasting industry audience figures and measures of particular systems coverage and speed of dissemination--or system effectiveness. In addition to two versions of the warning effectiveness standards, outdoor warning, EBS (crisis), EBS with CHAT-TV and a telephone warning system are used in a first test of the method. Use of the optimum mixture in programs for estimating increased survivors is also described.