Benchmarking: An International Journal, v. 24, n. 3, pp. 806-823. Abstract Purpose: The purpose of this paper is to measure the efficiency and the productive slacks of a sample of major UK auditing firms between 2005 and 2012. Design/methodology/approach: Using the bootstrapping technique as the cornerstone method, DEA estimates were computed, allowing the test for differences in the levels of efficiency and in the potentials for decreasing inputs and increasing outputs. Then, neural networks were combined with DEA estimates to model, with effective predictive ability, the drivers of auditing firms’ performance. Findings: The findings indicate an ambiguous impact of regulatory policies on efficiency levels. Results also indicate an eventual capacity shortfall, since the most important productive resource does not seem to cope with demand growth in the future. Originality/value: This paper aims to focus on efficiency, complementing previous research in the UK that focuses on productivity, and helps in establishing a decision-support system to slack evaluation.