This paper analyzes the relationship of fraud risk assessments to other risk assessments by auditors. The PCAOB (2005, 17) notes this is a problem area of current practice. The major findings are as follows. First, the study identifies the crucial role of benchmarks in forensic accounting to help differentiate between intentional and unintentional misstatements. Second, the paper shows the importance of not allowing the major categories of risks identified here from getting too high. This explains the need to set acceptable levels of these risks either by standard setters as a matter of broad policy, or by individual practitioners as part of the terms of specific engagements. Third, this paper clarifies how the fraud risk of SAS No. 99 relates to the traditional audit risk model framework.