This paper compares the photocatalytic and photoelectrocatalytic decolouration of dilute solutions of two dyes, one, methylene blue, cationic and one, reactive orange, anionic. For both methylene blue and reactive orange there is evidence that, as in photocatalytic decolouration, the initial stage in photoelectrocatalytic degradation is the destruction of the chromophore responsible for absorption in the visible region of the spectrum. For both dyes, both photoelectrocatalysis and photocatalysis were more efficient at pH > 9 than at pH 4. For the cationic dye, methylene blue, photoelectrocatalytic degradation was slower than photocatalytic degradation - indeed even though photocatalytic decolouration of 0.2 mM methylene blue solutions could be achieved photoelectrocatalytic degradation was negligible under the conditions used by us. However, photoelectrocatalytic decolouration of 0.1 mM methylene blue solution was achieved. The reduced efficiency of photoelectrocatalysis probably occurs because adsorption of the dye molecules on the TiO₂ surface is a necessary first step in photocatalysis and because the positive potential on the photanode reduces adsorption of the cationic methylene blue. For the anionic dye, reactive orange, photoelectrocatalytic degradation was more effective, under the conditions of our experiment. However, the rate of decolouration was approximately the same as that of photocatalytic degradation by a TiO₂ immobilized on a non-woven paper. The rate of both photocatalytic and photoelectrocatalytic degradation of the negatively charged Reactive Orange was faster at pH 10 than at pH 7, even though at the higher pH the surface of the TiO₂ may be expected to be more negatively charged.