This study examines whether the presentation order of audit evidence affects audit judgment and audit opinion decision and leads to recency effect. It also explores whether self-review can mitigate such bias. By manipulating response mode (end-of-sequence vs. step-by-step), presentation order of audit evidence (mitigating-contrast vs. contrast-mitigating), and debiasing mechanism (self-review vs. no self-review), and controlling the initial belief, an experiment with practicing auditors as the subjects was conducted. Subjects were asked to make judgments about a hypothetical client’s going concern likelihood and to make audit opinion choice based on the case material provided to them. Experimental data based on 78 auditors show that auditors’ going concern judgments are subject to recency bias and that self-review can effectively mitigate such bias. But, recency effect does not occur in the audit opinion decisions.