90 practicing auditors were given identical sets of financial statements except for the amount of contingent liability which was varied independently over three amounts of $50,000, $500,000 and $1 million. Based on this information, they were to indicate on a 7-point rating scale anchored by 1 (little likelihood) to 7 (great likehhood), their likelihood of issuing an unqualified, i.e., clean, report. 65 useable responses were received. In accordance with the theory of tolerance for ambiguity, it was hypothesized that auditors who were rated on the MacDonald AT-20 scale as being intolerant of ambiguity would have less preference for an unqualified audit report at higher amounts of contigent liability than auditors who were rated as being tolerant of ambiguity. A between-subjects analysis of variance showed that the auditors' tolerance for ambiguity interacted with different amounts of contingent liability to affect the likelihood of their issuing an unqualified report. Implications in terms of auditors' avoidance of litigation by the client are also discussed.