Understanding the head is one of the great challenges in the fields of comparative anatomy, developmental biology, and palaeontology of arthropods. Numerous conflicting views and interpretations are based on an enormous variety of descriptive and experimental approaches. The interpretation of the head influences views on phylogenetic relationships within the Arthropoda as well as outgroup relationships. Here, we review current hypotheses about head segmentation and the nature of head structures from various perspectives, which we try to combine to gain a deeper understanding of the arthropod head. Though discussion about arthropod heads shows some progress, unquestioned concepts (e.g., a presegmental acron) are still a source of bias. Several interpretations are no longer tenable based on recent results from comparative molecular developmental studies, improved morphological investigations, and new fossils. Current data indicate that the anterior arthropod head comprises three elements: the protocerebral/ocular region, the deutocerebral/antennal/cheliceral segment, and the tritocerebral/pedipalpal/second antennal/intercalary segment. The labrum and the mouth are part of the protocerebral/ocular region. Whether the labrum derives from a former pair of limbs remains an open question, but a majority of data support its broad homology across the Euarthropoda. From the alignment of head segments between onychophorans and euarthropods, we develop the concept of "primary" and "secondary antennae" in Recent and fossil arthropods, posit that "primary antennae" are retained in some fossil euarthropods below the crown group level, and propose that Trilobita are stem lineage representatives of the Mandibulata.