Objective To study the effect of sevoflurane on the early brain injury (EBI) after subarachnoid hemorrhage (SAH) through the focal death pathway mediated by the Nod‑like receptor associated protein 3 (NLRP3). Methods A total of 108 male C57BL/6 mice of 10−12 weeks were divided into three groups (36 mice in each group) according to the random number table method, namely the sham operation group (group A), the model group (group B) and the model+sevoflurane group (group C). The SAH model was made by the threading method in groups B and C, and the same operation was performed in group A, but the middle cerebral artery was not punctured. After establishing the SAH model, the mice in group C continued to inhale 3% sevoflurane and air for 1 h. The mouse in each group was evaluated by SAH grading and neurological function score, brain water content, and evans blue (EB) staining was used to calculate the amount of EB exudation to evaluate the permeability of blood brain barrier in mice. Western blot was used to detect matrix metalloprotein 9 (MMP‑9) and tight junction protein ［zonula occludens‑1 (ZO‑1), accludin and claudin‑5］ and inflammation‑related proteins NLRP3, apoptosis‑associated speck‑like protein containing CARD (ASC), cleaved caspase‑1, interleukin (IL)‑18, IL‑1β expression. Rreal‑time fluorescence quantitative polymerase chain reaction (PCR) detection of NLRP3 mRNA, caspase‑1 mRNA expression levels, detection of caspase‑1 activation. Terminal deoxynucleotidyl transferase‑mediated dUTP‑biotin nick end labeling (TUNEL) and NeuN double staining was used to detect neuronal pyroptosis. Results SAH bleeding grade and neurological function score showed that compared with group A, the SAH grade of group B was significantly increased (P<0.05), and compared with group B, the SAH grade of group C was significantly decreased (P<0.05). Compared with group C, the neurological score of group B was significantly reduced (P<0.05). Compared with group A, the brain water content and EB staining of group B increased significantly (P<0.05). Compared with group B, the brain water content and EB staining of group C decreased significantly (P<0.05). Compared with group A, the protein expression levels of ZO‑1, occludin and claudin‑5 of group B were significantly reduced (P<0.05), and the protein expression levels of NLRP3, ASC, cleaved caspase‑1, IL‑1β and IL‑18 of group B were significantly increased (P<0.05). Compared with group B, the expression levels of ZO‑1, occludin and claudin‑5 in group C were significantly increased (P<0.05), and the expression levels of NLRP3, ASC, cleaved caspase‑1, IL‑1β and IL‑18 were significantly decreased (P<0.05). Compared with group A, the expression level of MMP‑9 in group B was significantly increased. Compared with group B, the expression level of MMP‑9 in group C was significantly decreased (P<0.05). Real‑time quantitative PCR experiments showed that compared with group A, the mRNA levels of NLRP3 and caspase‑1 in group B were significantly increased; compared with group B, the mRNA levels of NLRP3 and caspase‑1 in group C were significantly decreased (P<0.05). Compared with group A, the activity of caspase‑1 in group B was greatly increased; compared with group B, the activity of caspase‑1 in group C was significantly decreased (P<0.05). TUNEL staining test showed that compared with group A, the number of TUNEL‑positive neurons in group B was significantly increased (P<0.05); compared with group B, the number of TUNEL‑positive neurons in group C was significantly decreased (P<0.05). Conclusions Sevoflurane can inhibit the NLRP3 inflammasome‑mediated pyroptosis after SAH and improve early brain injury after SAH.