When there is significant doubt about a firm's ability to continue as a going concern, professional standards require independent auditors to disclose the uncertainty in their report. This study assesses the influence of the independent auditor's going-concern evaluation by examining default following the release of the auditor's report. We use a proprietary sample maintained by the Portuguese Central Bank on 12,199 audit reports relating to approximately 2000 firms that are liable by law to have their accounts audited on an annual basis. Empirical estimation of a logit model controlling for accounting cash- flow-related and nonaccounting variables shows that the likelihood of default for firms that received going concern opinion is 2.792 times that of firms that received a clean opinion. Likelihood ratio tests for omitted variable also confirm the incremental predictive ability of going-concern opinion over and above accounting and nonaccounting variables for the estimation and hold-out samples. In the nondefaulting group, the average default rate is 6.05%, in the defaulting group it is 17.78%. The default rate for firms in the nondefaulting group that received a going-concern opinion is 9.92% and for firms that received a clean opinion it is 5.96%. In the defaulting group, the rate for firms that received a going-concern opinion is 35.49% and for firms that received a clean opinion it is 16.96%. Checks for robustness across different asset classes, age, industries, and regions indicate that firms that receive a going-concern opinion on average default more than those that receive a clean opinion.