This essay argues that the methodological and normative commitments of audit judgement research constrain its ability to address the expressive and persuasive dimensions of audit work. In contrast, a sociology of audit technique is proposed in which the "technicality" of audits is theorised as a product of process of closure rather than as a self-evident feature of audit practice. Accordingly, planning aids, such as the audit risk model, function more as a basis for representing the audit as a rational task than to guide procedures in the field. While these themes may appear to be counterintuitive, they emerge from a close reading of some of the audit judgement literature itself which is unable to develop the symbolic and justificational dimension of the audit process within the constraints of its theoretical framework. The sociology of audit technique inverts the cognitive assumptions of audit judgement research and critically recast the question of auditor expertise.