Objective To evaluate the predictive value of controlling nutritional status (CONUT) score and prognostic nutrition index (PNI) for 6⁃month prognosis of patients with severe traumatic brain injury (STBI). Methods A total of 100 patients with STBI into our hospital from January 2018 to June 2020 were retrospectively summarized. The modified Rankin Scale (mRS) was used to evaluate the 6⁃month clinical prognosis. The best cut⁃off value of CONUT score and PNI for poor prognosis was obtained by receiver operating curve (ROC curve). The relationship between CONUT score,PNI and clinical datas was analyzed, and univariate and multivariate Logistic regression analysis to screen the risk factors of prognosis. Results There were 53 patients (53.0%) of poor prognosis (mRS≥3). The best cut⁃off value of CONUT score was 4, and PNI was 44. The patients with CONUT score ≥4 or PNI <44 were older than CONUT score < 4, or PNI ≥44, Glasgow Coma Scale (GCS) score was less, cerebral hernia and pneumonia were more, serum albumin and total cholesterol were lower,too(P<0.05). Univariate regression analysis showed that age ≥50 years, GCS score < 6, cerebral hernia, pneumonia, albumin < 3.8 g/dl, total cholesterol<130 mg/dl, CONUT score ≥4 and PNI < 44 were associated with poor prognosis (P<0.05). Multivariate regression analysis showed that GCS score<6, pneumonia, CONUT score ≥4 and PNI <44 were independent risk factors for poor prognosis (P <0.05). Conclusion CONUT score and PNI before treatment have good predictive values for 6⁃month prognosis of STBI patients. CONUT score ≥ 4 and PNI < 44 are independent risk factors for poor prognosis.