The impact of sand on the efficiency of metallic iron (Fe-0) at discoloring a methylene blue (MB) solution in the presence of MnO2 was characterized. The MB initial concentration was 10 mg L-1. 22.0 mL of this solution was used for each test. Investigated systems were: (i) Fe-0 alone, (ii) MnO2 alone, (iii) sand alone, (iv) a sand/MnO2 mixture, and (v) Fe-0/sand/MnO2 mixtures with three different natural MnO2-bearing minerals. Tested material loadings varied from 0.0 to 45.0 g L-1. A commercial Fe-0 and three natural MnO2 materials (Manganite, Psilomelane and X-MnO2) were tested in shaken experiments at 75 rpm for 14 days. The documented delay of MB discoloration by Fe-0 in the presence of MnO2 was the evaluation criterion. This characteristic discoloration delay was observed for all tested materials at MnO2 loadings <6 g L-1. For higher MnO2 loadings, the Fe-0/Manganite/sand system was more efficient than the Fe-0/sand system. The largest extend of the delay of MB discoloration was observed in the Fe-0/X-MnO2/sand system. This observation delineated the crucial importance of characterizing the intrinsic reactivity of tested MnO2 materials for comparable results. This study highlighted the crucial importance of experimental conditions as major source of confusing! conflicting reports. (C) 2014 Elsevier B.V. All rights reserved.