Our aim was to study the mechanisms of brain histogenesis. As a model, we have used the role of winged helix transcription factor gene Foxb1 in the emergence of a very specific morphological trait of the diencephalon, the mammillary axonal complex. Foxb1 is expressed in a large hypothalamic neuronal group (the mammillary body), which gives origin to a major axonal bundle with branches to thalamus, tectum and tegmentum. We have generated mice carrying a targeted mutation of Foxb1 plus the tau-lacZ reporter. In these mutants, a subpopulation of dorsal thalamic ventricular cells (“thalamic palisade”) show abnormal persistence of Foxb1 transcriptional activity; the thalamic branch of the mammillary axonal complex is not able to grow past these cells and enter the thalamus. The other two branches of the mammillary axonal complex (to tectum and tegmentum) are unaffected by the mutation. Most of the neurons that originate the mammillothalamic axons suffer apoptosis after navigational failure. Analysis of chimeric brains with wild-type and Foxb1 mutant cells suggests that correct expression of Foxb1 in the thalamic palisade is sufficient to rescue the normal phenotype. Our results indicate that Foxb1 is essential for diencephalic histogenesis and that it exerts its effects by controlling access to the target by one particular axonal branch.