Continuous auditing increases the coverage and frequency of analysis of a firm's activities, and has been touted as a powerful fraud deterrence and detection technique, but we identify and examine a potential unintended consequence. When continuous auditing is accompanied by more timely notifications to auditees of exceptions to control rules, information is revealed about the system's capability to flag exceptions to control rules. Therefore, if a system has weak fraud-detection capability, early notification that the system did not detect a fraudulent transaction could actually increase an auditee's propensity to commit fraud. We examine whether the benefit of early notification depends on the fraud-detection capability of the organization's monitoring system (i.e., whether it is a strong or weak monitoring system). We use an experimental economics approach to address our research question. Consistent with expectations, we find that early and frequent notification of audit results is not always beneficial in deterring fraud, and that its benefit depends on whether the fraud-detection capability of the monitoring system is strong or weak. We do not find evidence of the predicted benefit of continuous notification reducing the incidence of fraud when the system is strong, but we do find an increase in participants' inclination to commit fraud when the system is weak. We discuss the implications of these findings for research and practice.