The prescriptive nature of the rule-based approach to anti-money laundering (AML) compliance and the exponential growth of suspicious activity reports resulting from the use of ‘tick-box’ method led to the adoption of the risk-based approach (RBA) to AML. However, the RBA suffers from a number of difficulties and this has resulted in its ineffective implementation. The difficulty in implementing the approach can mainly be traced back to the inappropriateness of the risk theoretical framework for AML. We argue that an uncertainty-based approach will provide a more fruitful basis for an effective AML regime. The paper, therefore, outlines how an uncertainty-based approach could be considered as an alternative to the existing RBA in order to improve the effectiveness of the AML compliance regime.