Previous experiments with mouse chimeras demonstrated that cellular competition for antigen-speci"c survival signals plays a crucial role in the maintenance of the naive B cell repertoire. Transgenic (Tg) B cell populations in these chimeras have a shortened lifespan and poor competitive abilities as compared to more diverse non-Tg populations in the same mice. We develop a mathematical model to investigate the mechanism of B cell competition. The model allows for various B cell clones, generated in the bone marrow, to go into the peripheral circulation, where they compete speci"cally for various ligands providing survival signals. In the model we also "nd the observed poor competitive abilities of the Tg repertoire. Investigating the nature of the competition in the model, we "nd that most of the competition is `intraspeci"ca occurring largely within the clone of truly Tg B cells, and within the repertoire of leaky Tg and non-Tg B cells. This is con"rmed by analysing a simpli"ed version of the model, which only allows for intraspeci"c competition, and resembles a simple ecological model with density-dependent death. The fact that our model accounts for the data, casts doubt on a previous interpretation of the same data arguing that more diverse repertoires outcompete repertoires of lower diversity. Here, we conclude that most of the data can be explained with intraspeci"c competition, and formulate an experimental prediction that allows one to distinguish between the previous interpretation of inter-speci"c competition between repertoires, and the current interpretation of intraspeci"c competition.