The existence of somatic, site-specific recombination in the central nervous system (CNS) has long been hypothesized but has been difficult to investigate experimentally. The finding that RAG-1, which is thought to encode a component of the site-specific recombination machinery of the immune system, is transcribed in the central nervous system (J.J.M. Chun et al., 1991, Cell 64:189-200), has renewed interest in this issue. Two groups (M. Kawaichi et al., 1991, J Biol Chem 266:18,376-18,394; M. Matsuoka et al., 1991, Science 254:81-86) have now reported the results of transgenic mouse experiments designed to determine whether cells of the CNS can perform a site-specific recombination reaction similar to that of lymphocytes. Despite extensive similarities in the design of the two experiments, they yielded discordant results and contradictory conclusions. An analysis of the two studies suggests some explanations for the discrepancies and leads us to two conclusions: first, that the CNS does not carry out the same somatic, site-specific recombination reaction as is found in the immune system and, second, that the question of whether other site-specific recombination processes occur in the brain remains open and largely unaddressed.