Objective To assess the effectiveness and safety of erythropoietin (EPO) for cancer-related malignant anemia without radiotherapy or chemotherapy. Methods Randomized controlled trials (RCTs) or quasi-randomized controlled trials (quasi-RCTs) involving erythropoietin in the treatment of cancer-related malignant anemia were searched and identified from PubMed (1966 to Sept. 2009), EMBASE (1974 to Sept. 2009), The Cochrane Library (Issue 3, 2009), CBM (1978 to Sept. 2009), CNKI (1994 to Sept. 2009), VIP (1989 to Sept. 2009). We also handsearched relevant journals. Data were extracted and evaluated by two reviewers independently with specially designed extraction form. We evaluated the quality of the included studies by the Cochrane Handbook 5.0 recommend standard and analyzed data by Cochrane Collaboration’s RevMan 5.0. Results We included twelve trials. The quality of the included studies was poor. The grade of ten studies was B, and the grade of two studies was C. Meta-analyses showed that there were significant differences between erythropoietin and blank in volume of blood transfusion [SMD= –0.66, 95%CI (–1.14, –0.17), P=0.008], number need to transfusion [OR=0.60, 95%CI (0.39, 0.92), P=0.02], and the change of hemoglobin after two-week therapy [SMD=2.40, 95%CI (0.29, 4.52), P=0.03]. Conclusion The current evidence shows that EPO significantly benefits cancer-related malignant anemia. Well-designed RCTs with a larger sample size, longer intervention and follow-up periods are still needed.