The timing, mechanism and force, and the likelihood or otherwise of an innocent explanation of inflicted head injury in early childhood was thrown into doubt by three papers by Geddes et al. published in 2001 and 2003. They challenged the generally accepted position that the forces involved in the generation of subdural and retinal haemorrhages in infants are always considerable, and proposed in its place the hypothesis that an infant might suffer a craniocervical injury that caused respiratory disturbance leading to global cerebral hypoxia. The resultant hypoxia, brain swelling, and rise in central venous pressure could cause subdural leakage of blood from intracranial vessels and retinal haemorrhages.