In this paper the morphological argument for the conditioning of Sanskrit aspiration and deaspiration is renewed in theoretically current terms, bringing forth new arguments and examining previously undiscussed major weaknesses in the purely phonological (autosegmental) argument. Relevant phonological, morphological, and lexical facts are outlined. A theory accounting for the non-phonological aspiration throwback alternations is presented, and it is argued that two recent purely phonological accounts of aspiration throwback have failed to make a convincing case. Additional general and specific arguments for the relevance of process morphology in Sanskrit aspiration throwback are presented. (MSE)