Abstract This article presents a Bayesian Belief Network (BBN) approach for socio‐technical system reliability assessment. A human error BBN model quantifies error influences arising from user knowledge, ability, and task environment, combined with factors describing the complexity of user action and user interface quality. System reliability evaluation is achieved by the System Reliability Analyser tool, which enables the iterative manipulation of the human error model according to high‐level scenarios. © 2003 Wiley Periodicals, Inc. Syst Eng 6: 210–223, 2003