<b><i>Purpose:</i></b> Both external beam radiation therapy (EBRT) and selective-internal radiation therapy (SIRT) are implemented to treat unresectable intrahepatic cholangiocarcinoma (iCCA). The present study aimed to evaluate the efficacy of EBRT and SIRT in managing iCCA through a systematic review and meta-analysis. <b><i>Methods:</i></b> PubMed and Cochrane database were queried to search for studies published from January 2000 toJune 2020 without language restrictions. Median survival time, overall survival, and radiological response were extracted. Secondary outcomes such as complication rates, predictors of survival, and downstage to surgery were pooled. Patient-level survival data were obtained to generate Kaplan-Meier survival graph. Pooled outcomes were analyzed with a random-effect model. <b><i>Results:</i></b> Twenty-nine and 20 studies including 732 and 443 patients from the SIRT and EBRT groups were included in the present study. From initial radiation treatment, the median survival time for patients who underwent SIRT and EBRT were 12.0 (95% confidence interval [CI]: 10.8–14.6) and 13.6 (95% CI: 11.1–16.0) months, respectively. As first-line therapy, the median survival time was 36.1 (95% CI: 20.6–39.5) months for SIRT and 11.0 (95% CI: 9.3–13.6) months for EBRT. Both radiation modalities were effective in downstaging initially unresectable iCCA to surgery (SIRT: 30.5%; EBRT: 18.3%). Patients in the SIRT group encountered more post-embolization abdominal pain (6.9 vs. 2.2%), ulcer (1.0 vs. 0.5%), nausea (1.6 vs. 0.7%), anorexia (5.9 vs. 0%), thrombocytopenia (7.3 vs. 6.0%), hyperbilirubinemia (5.2 vs. 2.1%), and hypoalbuminemia (13.2 vs. 3.3%), whereas EBRT was associated with higher rates of anemia (0.6 vs. 7.5%) and neutropenia (6.5 vs. 11.0%). <b><i>Conclusions:</i></b> Both EBRT and SIRT were safe and effective in treating unresectable iCCA. However, available evidence was highly heterogeneous regarding patient population, limiting fair comparison between 2 radiation modalities. Future high-quality comparative studies are warranted.