Neurosurgical procedures can result in brain injury by various means including direct trauma, hemorrhage, retractor stretch, and electrocautery. This surgically-induced brain injury (SBI) can cause post-operative complications such as brain edema. By creating a mouse model of SBI, we tested whether NADPH oxidase, an important reactive oxygen species producing enzyme, is involved in SBI using transgenic mice lacking gp91phox subunit of NADPH oxidase (gp91phox KO) and apocynin, a specific inhibitor of NADPH oxidase. Neurological function and brain edema were evaluated at 24 h post-SBI in gp91phox KO and wild-type littermates grouped into SBI and sham-surgery groups. Alternatively, mice were grouped into vehicle- and apocynin-treated (5 mg/kg, i.p. 30 min before SBI) groups. Oxidative stress indicated by lipid peroxidation (LPO) was measured at 3 and 24 h post-SBI. The gp91phox KO mice, but not the apocynin-treated mice showed significantly improved neurological scores. Brain edema was observed in both gp91phox KO and wild-type groups after SBI; however, there was no significant difference between these two groups. Brain edema was also not affected by apocynin-pretreatment. LPO levels were significantly higher in SBI group in both gp91phox KO and wild-type groups as compared to sham group. A trend, although without statistical significance, was noted towards attenuation of LPO in the gp91phox KO animals as compared to wild-type group. LPO levels were significantly attenuated at 3 h post-SBI by apocynin-pretreatment but not at 24 h post-SBI. These results suggest that chronic and acute inhibition of NADPH oxidase activity does not reduce brain edema after SBI. Long-term inhibition of NADPH oxidase, however improves neurological functions after SBI.