A ketogenic diet (KD) improves cellular metabolism and functional recovery after moderate-to-severe traumatic brain injury. Here, we evaluated the changes of neurochemical metabolites after KD therapy for repetitive mild traumatic brain injury (rmTBI) and its possible role in neurodegeneration. Postnatal day 35 rats were randomly divided into 3 groups: sham, control, and KD groups. Rats in control and KD groups were given 3 rmTBI by a fluid percussion traumatic brain injury device 24 hours apart. All rats were killed at 7 days after the last injury. The ipsilateral cortex were analyzed with hematoxylin and eosin staining; beta-hydroxybutyrate was measured; conventional magnetic resonance imaging and the dry-wet weight method were used to assess the brain edema; changes of neurochemical metabolites were assessed using the ratio of N-acetylaspartate (NAA)/creatine (Cr), choline compound (Cho)/Cr, and NAA/Cho with magnetic resonance spectroscopy; the effect of KD therapy on neurodegeneration was evaluated with double immunofluorescence staining of Iba-1/beclin-1; behavioral outcome was assessed with beam walk/beam balance tests. KD significantly elevated beta-hydroxybutyrate levels, and there was no brain edema associated with rmTBI and KD therapy; behavioral assessment showed KD therapy significantly improved motor performance; magnetic resonance spectroscopy showed that rmTBI reduced the ratio of NAA/Cr and had no effect on the ratios of Cho/Cr and NAA/Cho whereas KD increased the ratio of NAA/Cr; double immunofluorescence staining showed KD therapy could significantly decrease microglial beclin-1 expression in the ipsilateral cortex. These results suggest the effect of KD on metabolic status and its possible role in preventing neurodegeneration in adolescent rats after rmTBI.