Cationic porphyrins have been shown to photoinduce the direct inactivation of Gram-positive (G+) and Gram-negative (G-) bacteria, thereby differing from anionic or neutral porphyrins which can photosensitize the G- bacteria only after permeabilization of their outer membrane. The present data show that the differences between these positively and negatively charged porphyrins are not related by a difference in the intrinsic photosensitizing efficiency, as determined by the photo-oxidation of model substrates or the yield of 1O2 generation; moreover, there are only minor differences in the quantum yield of porphyrin photobleaching. Rather, it appears that the positive charge promotes an electrostatic binding of the porphyrin to the outer cell surface inducing an initial limited damage which favours the penetration of the photosensitizer. Actually, the overall photoprocess is inhibited by the preincorporation of the porphyrin into liposomes, while it is enhanced by using amphiphilic dicationic porphyrins which bind to endocellular sites in larger amounts and in a more stable form.