Mathematical models have been used to study different aspects of the germinal centre reaction, in particular, affinity maturation of antibodies and the hypothesis of recycling. So far, interpretation of several theoretical and experimental results has pointed to the existence of recycling. However, theoretical models have seldom been compared with experimental data from specific immune responses and the potential relevance of recycling in the germinal centre is still an open problem. In this article, we propose a model without recycling that takes into account selection mechanisms that were previously uncovered experimentally. We apply the model to several experimental systems that use different Ag and compare the results with experimental data of affinity maturation whenever available. The results obtained for a primary immune response to the hapten (4-hydroxy-3-nitrophenyl)-acetyl show that recycling is not a necessary mechanism to achieve the level of affinity maturation observed in germinal centre reactions. Similar levels of affinity maturation are obtained for other responses, although for antibodies involving several affinity-enhancing mutations the affinity maturation obtained with the model is much lower. Interpretation of these results and consequences towards the concept of recycling are discussed.