TARGET: To observe the effect on hemoglobin (Hb) variability with different treatment frequencies of intravenous iron in maintenance hemodialysis (MHD) patients. METHODS: There were 41 MHD patients enrolled in the cohort. The patients were separated into two groups randomly. The baseline data were collected after oral iron agents for 1 month (wash out). There were two methods of intravenous iron administration, either continuous or intermittent. For continuous administration (CA), 100 mg sucrose iron agents were given during every HD session with total dose of 1000 mg. For intermittent administration (IA), 100 mg sucrose iron agents were given once every week with the same total dose of 1000 mg. The protocol were designed to do a follow-up of 7 months, which included two to three-month steps of different administrations of intravenous iron and 1 month of wash-out period by oral iron agents between two steps, respectively. Patients in Group one (G1) administrated iron agents by CA at the first step and IA at the second step. Patients in Group two (G2) did IA and then transfer to CA. The clinical parameters including Hb, serum ferritin (SF), transferrin saturation (TAST), and doses of recombined human erythropoietin (rHuEPO) were detected and recorded during follow-up period. The standard deviation of Hb (Hb-SD) and coefficient variation of Hb (Hb-CV) were calculated. The baseline data were compared between two groups. The parameters on month 3 and 7 were compared with those on month 0 in two groups, respectively. The effects of both CA and IA on Hb-SD and Hb-CV were compared by two-stage cross-comparison general linear model (GLM) method. RESULTS: There were 34 patients (82.9%) completed the trail. The amounts of cases, rates of gender and primary diseases, the mean value of age, Hb, SF, TSAT, and doses of rHuEPO on baseline in two groups were similar (p > 0.05). The SF levels in two groups increased significantly during follow-up period, which were 235.4 ± 51.8 ng/ml on month 0, 362.4 ± 140.2 ng/ml on month 3, and 315.0 ± 97.73 ng/ml on month 7 in G1 (p < 0.01), and 250.5 ± 37.8 ng/ml, 332.2 ± 118.9 ng/ml, and 347.4 ± 124.3 ng/ml in G2 (p < 0.01), respectively. Compared to CA, IA could decline the Hb-SD (5.93 ± 3.97 g/l vs. 7.36 ± 3.81 g/l, F = 4.377, p = 0.044) and Hb-CV (0.054 ± 0.035 vs. 0.069 ± 0.030, F = 7.042, p = 0.012) significantly. The mean levels of Hb and doses of rHuEPO were similar between CA and IA. CONCLUSION: The administration of intravenous iron by CA or IA has the similar effects on iron supplement and anemia treatment. However, IA may be more benefit to Hb variability than CA in MHD patients.