We investigated brain development in the horseshoe crab Limulus polyphemus and several other arthropods via immunocytochemical methods, i.e. antibody stainings against acetylated alpha-tubulin and synapsin. According to the traditional view, the first appendage-bearing segment in chelicerates (the chelicerae) is not homologous to the first appendage-bearing segment of mandibulates (first antenna, deutocerebrum) but to the segment of the second antenna (tritocerebrum) or the intercalary segment in hexapods and myriapods. Accordingly, the segment of the deutocerebrum in chelicerates would be completely reduced. The main arguments for this view are: (1) the postoral origin of the cheliceral ganglion, (2) a poststomodaeal commissure, and (3) a connection of the cheliceral ganglion to the stomatogastric system. Our data show that these arguments are not convincing. During the development of horseshoe crabs there is no evidence for a former additional segment in front of the chelicerae. Instead, comparison of the brain structure (neuropil ring) between chelicerates, crustaceans and insects shows remarkable similarities. Furthermore, the cheliceral commissure in horseshoe crabs runs mainly praestomodaeal, which would be unique for a tritocerebral commissure. An unbiased view of the developing nervous system in the "head" of chelicerates, crustaceans and insects leads to a homologisation of the cheliceral segment and that of the (first) antenna (= deutocerebrum) of mandibulates that is also congruous to the interpretation of the Hox gene expression patterns. Thus, our data provide morphological evidence for the existence of a chelicerate deutocerebrum.