The double S shape of the vertebral column is one of the most important evolutionary adaptations to human bipedal locomotion, providing an optimal compromise between stability and mobility. It is commonly believed that a six element long lumbar spine facilitated the critical adoption of lumbar lordosis in early hominins, which contrasts with five lumbars in modern humans and four in chimpanzees and gorillas. This is mainly based on the juvenile Homo erectus skeleton KNM-WT 15000 from Nariokotome, Kenya. Yet, the biomechanical advantage of a long lumbar spine is speculative. Here we present new vertebral and rib fragments of KNM-WT 15000. They demonstrate that the sixth to the last presacral vertebra possesses rib facets and therefore indicate the presence of only five lumbar and twelve thoracic segments, as is characteristic of modern humans. Moreover, they show that no additional element was located between the sixth to the last presacral vertebra and Th11 as suggested in the original description. The transition from thoracic to lumbar type orientation of the facet joints that takes place at Th11 is thus at the same segment as in over 40% of modern humans, suggesting an identical lumbar mobility and capacity for lordosis. Taken together, KNM-WT 15000 had one vertebra less than previously thought irrespective of whether rib-free lumbar vertebrae or vertebrae that bear lumbar-like articular processes are counted. Furthermore, the new rib fragments imply a rearrangement of the ribs that results in a symmetrical rib cage. This challenges previous claims for idiopathic or congenital scoliosis. We conclude that the bauplan of the hominin axial skeleton is more conservative than previously thought.