Existing reliability assessment methods are not consistent, and the approaches used produce inefficiencies in resource allocation. In this paper, a comprehensive new probabilistic reliability assessment framework is proposed and new reliability assessment methods and approaches are developed. Novel approaches dealing with probabilistic voltage stability analysis, angle stability, power balance and integrated reliability planning for bulk power systems are also developed in this work. Seven performance states have been proposed to define the power system conditions. System issues and a Markov model defining these states are also presented. New probabilistic voltage and angle stability assessment approaches are developed, including new voltage stability measures (e-indices) based on machine and system reactive margins. Application consideration of the new reliability framework is also presented. All model developments were tested with a real electric power system of the western U.S.A.