Scaphoid fracture union depends upon early diagnosis and adequate management and is considered to be adversely influenced by proximal scaphoid avascularity and carpal instability. There is, however, controversy on various aspects in the diagnosis and management of this injury. Studies investigating some of these aspects are reported in this thesis. With regard to the initial diagnosis of this fracture this thesis demonstrated that the fracture was usually visible on the 'Ziter' or the posterio-anterior radiographic view while radiographic evaluation of soft tissue swelling was of less value than that suggested by current literature. Moreover, the second week radiograph did not provide reliable evidence of a fracture in those suspected on clinical examination alone. Normal lines in an intact scaphoid could be mis-interpreted as a fracture. With regard to management, the fracture healed equally well in a Colles' cast: thumb immobilisation did not appear to be neccessary. The incidence of non-union was around 12% and most were symptomatic. A further 25%, in whom the fracture site was easily identified but appeared to have healed, formed a distinct clinical group and many had symptoms. The reliability and reproducibility of radiographic signs of union were poor. Therefore, a real time ultrasonographic method of demonstrating movement at the scaphoid fracture site was developed. Proximal avascularity was more common than that suggested by radiographic appearances alone while carpal instability was uncommon within two years of injury. Based on these observations a rational approach to management of scaphoid fractures was proposed.