Iron can be anodically electropolished in non-aqueous solution of sulphamic acid (2·9 N HSO3NH2 in HCONH2) at relatively low current density. The galvanostatic polarization curves determined at steady states at 20°C and 40°C show apparently two discontinuous curves, probably caused by passivation. Periodic phenomena can be frequently observed near the transition region.By measurement of electrode potentials in potentiostatic polarization and decay curves, as well as by analytical methods, the mechanism of electropolishing is confirmed to be due to passivation occurring over the anode surface.The Flade potential, and its pH-dependence, which characterizes the transition potential between the active and passive states in aqueous systems, can also be obtained by discharge curves of the passivated iron anode. It is, however, 0·38 V at pH = 0 (the standard state), being rather less noble than 0·56 V which Flade and others have obtained in aqueous solution.By analysis of the anolyte, the basic anode reaction on a Pt surface is considered to be2HCONH− → HCONH2 + HCNO + 2e; E0 = 0·38 + RT/2F ln aHCNO The standard electrode potential of this reaction is found experimentally to be 0·38 V in 1 N HSO3NH2 solution, coincident with thermodynamic calculation. This fact strongly suggests that generation of oxygen cannot occur since HCONH− ion can more easily discharge above 0·38 V and that the original product produced on the anode is HCNO or related substances. Even by adding a slight amount of water intentionally, no shift of the Flade potential is found, the potential being only somewhat stabilized.To account for the cause of passivation, three assumptions, the formation of an oxide film, the formation of an unknown organic film containing formamide, and the formation of an HCNO gas layer are proposed and it is suggested that either of the latter two is the most probable.