PURPOSE: To evaluate the safety (risk of infection) and efficacy (transfusion requirements, changes in haemoglobin (Hb)) of iron therapy in adult intensive care unit (ICU) patients. MATERIALS AND METHODS: We systematically searched seven databases for all relevant studies until January 2018 and included randomized (RCT) studies comparing iron, by any route, with placebo/no iron. RESULTS: 805 participants from 6 RCTs were included. Iron therapy, by any route, did not decrease the risk of requirement for a red blood cell (RBC) transfusion (Risk ratio (RR) 0.91, 95% CI 0.80 to 1.04, p = 0.15) or mean number of RBCs transfused per participant (mean difference (MD) -0.30, 95% CI -0.68 to 0.07, p = 0.15). Iron therapy did increase mean Hb concentration (MD 0.31 g/dL, 95% CI 0.04 to 0.59, p = 0.03). There was no difference in infection (RR 0.95, 95% CI 0.79 to 1.19, p = 0.44). Trial Sequential Analysis suggests that the required participant numbers to detect or reject a clinically important effect of iron therapy on transfusion requirements or infection in ICU patients has not yet been reached. CONCLUSION: Iron therapy results in a modest increase in Hb. The current evidence is inadequate to exclude an important effect on transfusion requirements or infection.