The most common malignant central nervous system tumor is glioblastoma multiforme (GBM). Cytokine-induced killer (CIK) cell therapy is a promising type of adoptive cell immunotherapy for various cancers. We previously conducted a randomized clinical trial on CIK cell therapy in patients with GBM. The aim of this study was to evaluate the efficacy of CIK immunotherapy for patients with pathologically pure GBM, using data from our previous randomized clinical trial. The difference between overall survival (OS) and progression-free survival (PFS) according to CIK immunotherapy was analyzed using the Kaplan-Meier method. Hazard ratios were calculated using univariate and multivariate Cox regression analyses to determine whether CIK cell immunotherapy was independently associated with higher OS and PFS in patients with pure GBM. A total of 156 eligible patients were included in the modified intention-to-treat (mITT) population. We confirmed that 125 (80.1%) GBM samples were pure GBM tumors without the presence of other types of tumors. For patients with pure GBM, Kaplan-Meier analysis showed no significant difference in OS between the CIK cell treatment and control groups. However, multivariate Cox regression demonstrated CIK cell immunotherapy as an independent predictor of greater OS (hazard ratio, 0.59; 95% CI, 0.36-0.97; p = 0.038) and PFS (hazard ratio, 0.55; 95% CI, 0.36-0.84; p = 0.001) in patients with pathologically pure GBM in the mITT population. This study showed that CIK cell immunotherapy combined with conventional temozolomide chemoradiotherapy could prolong OS and PFS in patients with newly diagnosed pathologically pure GBM, with no significant adverse events related to treatment. However, unlike the results of multivariate Cox analysis, no statistical significance of CIK cell immunotherapy in OS in Kaplan-Meier analysis raises a question. Further studies are required to validate these results.