Background: The purpose of this study was to investigate the difference between two nailing approaches of intramedullary screw fixation, the retrograde nailing versus the anterograde nailing, on the radiological and clinical outcomes in patients with clavicle shaft fractures. Methods: From April 2002 to August 2014, we enrolled a total of 22 patients with clavicle shaft fractures to participate in this study. Twelve patients received retrograde intramedullary nailing and 10 received anterograde nailing. The average duration of follow-up was 12 months. In all the patients, we took follow-up radiographs of the anteroposterior and the axial views to assess the postoperative radiological outcomes. We measured the visual analogue scale (VAS) score, American Shoulder and Elbow Surgeons (ASES) score, and the range of motion (ROM). Results: Clinically, we did not find a statistically significant difference in the retrograde group and the anterograde group in terms of the duration to bone union, the VAS score the ASES score and the ROMs. Radiologically, we found that the difference in the clavicle shortening of the affected arm and the unaffected arm did not show a statistically significant difference at the immediate postoperative assessment. we found that the difference in the clavicle shortening of the affected arm between the immediate postoperative and the final follow-up value did not show a statistically significant difference. Conclusions: We found that both the retrograde nailing and the anterograde nailing gave a favorable outcome for clavicle shaft fractures. Although we saw evidence of clavicle shortening after intramedullary screw fixation, this was not a factor that influenced clinical outcome.