The paper (I) outlines the nature and extent of the dissatisfaction with legal research instruction and demonstrates that the problem predates computer-assisted legal research, (II) presents the history of the debate (focusing on a heated exchange between advocates of a "process-oriented" approach and proponents of the traditional, "bibliographic" methods), and (III) presents the requisite elements of a satisfactory pedagogical model, discussing various issues surrounding each of these elements.In part III, the paper proposes that a complete pedagogical model requires (A) an identifiable and fully understood objective in teaching legal research (which objective must distinguish between the kinds of research done by attorneys, scholars, and librarians), (B) a theory and understanding of the nature of legal source materials, (C) a theory of mathetics, or the nature of students and how they learn (with emphasis upon the provision of conceptual models for internalizing research techniques), and (D) a methodology consistent with the previous elements. Besides proposing the elements of the pedagogical model, this paper will explore the subtle issues surrounding each element (including examples of frameworks selected by the author), concluding that an appropriate pedagogical model has to be designed based upon the particular circumstances and needs of each law school.