The galvanostatic intemittent titration technique (GITT) has been used as a standard method for evaluating chemical diffusion coefficients in electrode materials in the last three decades. It will now be demonstrated that these chemical diffusion coefficients evaluated with GITT are spurious as any reaction kinetics is neglected in the GITT theory. The neglect of the reaction kinetics leads to a spurious potential dependence of the GITT diffusion coefficients with minima at those potentials where the slow scan rate cyclic voltammogram or differential capacity plot exhibits peaks even in case where the true chemical diffusion coefficient is constant. This will be demonstrated by the evaluation of GITT diffusion coefficients from numerically generated GITT experiments calculated with a constant chemical diffusion coefficient on the example of a spinel-type LiMn2O4 electrode. © 2004 Elsevier Ltd. All rights reserved.