PurposeThis study aims to address the following four research questions: first, whether auditors report critical audit matters (CAMs) to shield themselves against possible litigation; second, whether reporting quality affects auditors’ propensity to report CAMs; third, whether auditors’ tenure length – reflecting familiarity with clients’ financial reporting – affects their likelihood to report CAMs; and fourth, whether auditors’ conservatism increases the likelihood of CAMs reporting.Design/methodology/approachData are manually collected from audit reports including CAMs in 10-K, then financial data are collected from the Capital IQ database, and market data are collected from the CRSP database. Using propensity score matching, the initial sample of companies with CAMs is matched with companies without reported CAMs. Performance adjusted discretionary accruals, real earnings management proxy, Khan and Watts’ (2009) C-score, propensity to issue a going concern opinion, Dechow et al.’s (2011) F-Score, Rogers and Stocken’s (2005) model and Houston et al.’s (2010) model are used to measure reporting quality, auditor conservatism, misstatement risk and litigation risk, respectively.FindingsThe results do not show that auditors report CAMs opportunistically to shield themselves from litigation risk. However, the results do suggest that auditors have a greater tendency to report CAMs when reporting quality is low and when they are more conservative. On the other hand, they have less tendency to report CAMs in their first year of engagement.Research limitations/implicationsThe findings of this study have important implications for the auditor behavior literature as it shows that, when it comes to reporting CAMs, auditors actually behave objectively and do not report in a trite way. This study also provides early archival evidence on a standard that relates to the first major change to the auditor’s report in decades. To the best of the author’s knowledge, it is the first to provide evidence on the association between auditor conservatism and auditors tendency to report CAMs and the first to triangulate prior research on auditor litigation risk by providing the first archival evidence on the auditors “litigation-shielding” concern.Practical implicationsThis study examines whether auditors attempt to meet the stated objective of reporting CAMs by signaling information about reporting quality. This study demonstrates that reporting CAMs is not a “boilerplate” communication. This study has implications for standards setters, as it shows that CAMs are reported in a way consistent with the objectives of the new standard, namely, via signaling information in the audit report on the quality of the financial statements.Originality/valueIn terms of originality, this paper uses a manually collected sample and, to the best of the author’s knowledge, is the first to focus on auditor’s behavior rather than on investors or clients reactions to CAMs. Also, this paper addresses a recently issued standard using US data and archival approach, rather than experimental. This paper also provides relevant evidence related to concerns raised earlier but were not empirically examined, such as reporting CAMS as “boilerplate” expectations. This paper provides new evidence on the auditors’ behavior with regard to litigation risk.