Early theories of tax enforcement suggest that higher auditing levels lead to higher levels of tax compliance (Allingham and Sandmo, 1972; Srinivasan, 1973). During the last decades, however, empirical research has provided mixed evidence for this association (Beer et al., 2015; Gangl et al., 2014; Iyer et al., 2010; Slemrod et al., 2001; Spicer and Thomas, 1982). As a result, subsequent theoretical work has examined why auditing can have positive, negative, or null effects on tax compliance (Alm, 2012; Devos, 2014; Feld and Frey, 2007; Kirchler et al., 2008; Lederman, 2016; Torgler, 2002). 1 Despite these analyses and several cross-country examinations (e.g., Frey and Torgler, 2007; Fuest and Riedel, 2009; Richardson, 2006; Tsakumis et al., 2007), little is known about the association between auditing and tax compliance at the country-level of analysis – the level at which tax variables of great significance are predominantly set (e.g., central government tax rates, auditing procedures, penalties for noncompliance). This study aims at filling this gap, and responds to recent calls from multilateral organizations for research on the effectiveness of audit strategies employed in practice (IMF, 2015; OECD, 2010). Building on behavioral theories of tax compliance, we hypothesize that the association between auditing and tax compliance is non-linear. Our expectation is that compliance increases until a certain auditing level is reached, and decreases beyond that level (i.e., an elevated auditing level backfires). Put differently, we expect a U-shaped association between the auditing level and noncompliance. The rationale is that, while holding everything else constant, an elevated auditing level may signal distrust in taxpayers and lead to the perception that the tax authority and its enforcement actions are excessive and unfair (Feld and Frey, 2002; Kirchler et al., 2008; Lawsky, 2008; Mooijman et al., in press; Wahl et al., 2010; Wenzel, 2003). Perceived unfairness and feelings of distrust may in turn lead to reduced compliance (Hofmann et al., 2008; Kirchler, 2007; Murphy, 2005; Torgler, 2007). This goes in line with prior research suggesting that control in general, and crime deterrence in particular, can backfire when these are indicative of distrust and unfairness (Falk and Kosfeld, 2006; Kessler and Leider, 2016; Schildberg-H¨orisch and Strassmair, 2012). Using publicly available country-level data from 2003–۲۰۱۴, we examine the association between the auditing level, which is measured as the number of verification actions taken by tax authorities per 100 taxpayers in each country (OECD, 2017a),  and business executives’ perception of tax evasion (IMD, 2017a; World Bank, 2017b).