Ever since the seminal papers of Keith and Schultz, hominoid primate ribcages have been described as either "funnel-" or "barrel-shaped." Following this dichotomic typology, it is currently held that Homo sapiens and hylobatids (gibbons and siamangs) share a barrel-shaped ribcage and that they are more similar to each other than to the funnel-shaped thoraces of great apes (Gorilla, Pan, and Pongo). Other researchers hypothesized that thoracic width and the invagination of the thoracic spine into the thorax are related to allometry. However, analyses that take into account the complex three-dimensional (3D) shape of the ribcage are lacking. Here, we address hypotheses about thorax shape and evolution using 3D morphometrics of thoraces in anatomical connection obtained by computed tomography scans of 23 hominoid cadavers and 10 humans and examining thorax compartments composed of seven ribs (1-7 thorax) and of 11 ribs (1-11 thorax). In the 1-7 thorax analyses, the human thorax is uniquely flat because of torsion of the upper and central ribs, differing from all non-human hominoids including hylobatids. In the 1-11 thorax analyses, humans are markedly different from African great apes, with hylobatids and orangutans intermediate. In full shape space analyses, affinities between orangutans and humans on the one hand and between hylobatids and African great apes on the other are evident. Therefore, we reject the hypothesis that humans and hylobatids bear any special affinities in overall 3D thorax shape to each other. We find that larger thoraces are wider and flatter, with a more invaginated spine, supporting the allometric hypothesis. Hominoid thorax variation shows complex interactions between allometry, rib curves, torsion, and declination, and the morphology of the costo-vertebral joint and the thoracic vertebral column. When considering functional specializations alongside phylogenetic relationships, an overly simplistic dichotomy between funnel-shaped and barrel-shaped thoraces is not supported.