No fruitful discussion of the logic of religious belief can afford to ignore George Mavrodes' classification of propositional concepts, i.e. concepts predicable of propositions singly (p) or in sets (S), as an analytical tool for pinning down the ‘person-oriented’ and ‘content-oriented’ factors in such ‘epistemic activities’ as religious proving, experiencing, and verifying. Mavrodes shows in particular that the formal model of logical soundness, i.e. valid form and true premises, has but limited application to proving, experiencing, and verifying as ways of giving good reasons for religious belief. In this essay, I will show how Mavrodes' analysis of the logic of religious belief can be extended by substituting a procedural model of logical soundness for the truth-functionally formal model that he adopts. I will show further why it is desirable and, in certain instances, vital to extend his analysis in this manner. Along the way, I will explain why the extension that I am advocating augments Mavrodes' original analysis without substantially altering it or sacrificing any of its original analytical virtues.