This dissertation analyses the research funding resource allocation mechanism (the Research Assessment Exercise) in England to assess its viability as a resource allocation tool and a performance control measure, to form a view on both the internal consistency of the exercise and to explore possible unintended consequences. Case study interviews were carried out with university administrators to investigate the institutional impact. The academics' behaviour was researched by a questionnaire survey. A survey of journal editors was also carried out. Logistic regression was applied to the survey of academics to analyse the data. The RAE has resulted in a "publication culture", where academics are concentrating on research that produces early publishable results and a tendency to publish as many papers, as possible, from the same research project. The impact of the RAE on academics was not independent of their characteristics. The level of self-assessed research activity was a significant predictor variable. The 'middle-tier' academics were the most influenced by the RAE "four-paper" effect. Overall, the RAE lacked coherence and consistency as a resource allocation methodology, and had unintended consequences as a performance measure.