Screws with different levels of compression force are available for scaphoid fixation and it is known that the Acutrak screw generates greater compression than the Herbert screw. We retrospectively compared two types of headless compression screw for their effectiveness in the repair of scaphoid nonunion. Twenty-nine cases of proximal scaphoid nonunion were surgically treated with non-vascularised bone graft: the Acutrak screw was used in 17 patients and the cannulated Herbert screw in 12 patients. Wrist range of motion, Mayo wrist score, grip strength and QuickDASH scores were indicators used for the functional evaluation. Radiographic findings were assessed for consolidation of nonunion and signs of arthrosis. The mean follow-up time was 49.2 months (range 12-96). Statistically, there was no significant difference between the Acutrak and Herbert screw types in terms of functional evaluation and time required for consolidation. Greater compression did not influence the functional outcome, consolidation rate or time to consolidation. The need for greater compression in the treatment of proximal scaphoid nonunions is thus questionable because it may increase the risk of proximal fragment communition.