The weight-drop model is used widely to replicate closed-head injuries in mice; however, the histopathological and functional outcomes may vary significantly between laboratories. Because skull fractures are reported to occur in this model, we aimed to evaluate whether these breaks may influence the variability of the weight-drop (WD) model. Male Swiss Webster mice underwent WD injury with either a 2 or 5 mm cone tip, and behavior was assessed at 2 h and 24 h thereafter using the neurological severity score. The expression of interleukin (IL)-6, IL-1β, tumor necrosis factor-α, matrix metalloproteinase-9, and tissue inhibitor of metalloproteinase-1 genes was measured at 12 h and 1, 3, and 14 days after injury. Before the injury, micro-computed tomography (micro-CT) was performed to quantify skull thickness at the impact site. With a conventional tip diameter of 2 mm, 33% of mice showed fractures of the parietal bone; the 5 mm tip produced only 10% fractures. Compared with mice without fractures, mice with fractures had a severity-dependent worse functional outcome and a more pronounced upregulation of inflammatory genes in the brain. Older mice were associated with thicker parietal bones and were less prone to skull fractures. In addition, mice that underwent traumatic brain injury (TBI) with skull fracture had macroscopic brain damage because of skull depression. Skull fractures explain a considerable proportion of the variability observed in the WD model in mice-i.e., mice with skull fractures have a much stronger inflammatory response than do mice without fractures. Using older mice with thicker skull bones and an impact cone with a larger diameter reduces the rate of skull fractures and the variability in this very useful closed-head TBI model.