Laboratory experiments and modeling studies were performed to investigate the redox transformations of arsenic and iron in water treatment sludge during aging, and to evaluate the impact of those transformations on the leachability of arsenic determined with the U.S. EPA toxicity characteristic leaching procedure (TCLP). When the backwash suspension samples collected from a California surface water treatment plant were aged in closed containers for a few weeks, soluble arsenic increased from less than 5 microg/L to as high as 700 microg/L and then decreased dramatically because of biotic reduction of arsenate [As(V)], ferric oxyhydroxide, and sulfate. The experimental results and the thermodynamic models showed that arsenic mobility can be divided into three redox zones: (a) an adsorption zone at pe > 0, which is characterized by strong adsorption of As(V) on ferric oxyhydroxide; (b) a mobilization (transition) zone at -4.0 < pe < 0, where arsenic is released because of reduction of ferric oxyhydroxide to ferrous iron and As(V) to arsenite [As(III)]; and (c) a reductive fixation zone at pe < -4.0, where arsenic is immobilized by pyrite and other reduced solid phases. The TCLP substantially underestimated the leachability of arsenic in the anoxic sludge collected from sludge ponds because of the oxidation of Fe(II) and As(III) by oxygen. The leaching test should be performed in zero-headspace vessels or under nitrogen to minimize the transformations of the redox-sensitive chemical species.