To evaluate the neuroprotection exerted by ketosis against acute damage of the mammalian central nervous system (CNS). Search engines were interrogated to identify experimental studies comparing the mitigating effect of ketosis (intervention) vs non‐ketosis (control) on acute CNS damage. Primary endpoint was a reduction in mortality. Secondary endpoints were a reduction in neuronal damage and dysfunction, and an “aggregated advantage” (composite of all primary and secondary endpoints). Hedge’s g was the effect measure. Subgroup analyses evaluated the modulatory effect of age, insult type, and injury site. Meta‐regression evaluated timing, type, and magnitude of intervention as predictors of neuroprotection. The selected publications were 49 experimental murine studies (period 1979‐2020). The intervention reduced mortality (g 2.45, SE 0.48, p <0.01), neuronal damage (g 1.96, SE 0.23, p <0.01) and dysfunction (g 0.99, SE 0.10, p <0.01). Reduction of mortality was particularly pronounced in the adult subgroup (g 2.71, SE 0.57, p <0.01). The aggregated advantage of ketosis was stronger in the pediatric (g 3.98, SE 0.71, p <0.01), brain (g 1.96, SE 0.18, p <0.01), and ischemic insult (g 2.20, SE 0.23, p <0.01) subgroups. Only the magnitude of intervention was a predictor of neuroprotection (g 0.07, SE 0.03, p 0.01 per every mmol/L increase in ketone levels). Ketosis exerts a potent neuroprotection against acute damage to the mammalian CNS in terms of reduction of mortality, of neuronal damage and dysfunction. Hematic levels of ketones are directly proportional to the effect size of neuroprotection.