To find out if there are changes in membrane potential during bacterial chemotaxis, we measured the membrane potential of Escherichia coli indirectly by use of the permeating, lipid-soluble cation triphenylmethylphosphonium. Addition of attractants or repellents to the bacteria brought about a hyperpolarizing peak (as well as additional, later changes in membrane potential). This peak was shown to be a part of the chemotactic mechanism based on the following evidence: (i) All attractants and repellents tested gave this peak while chemotactically inert chemicals did not. (ii) Mutants lacking galactose taxis failed to give the peak with galactose but did with another attractant and with repellents. (iii) Methionine, required for chemotaxis, is also required for production of this peak. (iv) A mutant in a control gene )flaI), unable to synthesize flagella and cytoplasmic membrane proteins related to motility and chemotaxis, failed to give the peak. (v) Paralyzed (mot) mutants gave little or none of the peak. Generally nonchemotactic (che) mutants, on the other hand, did give this peak. Very likely there are ion fluxes that bring about this change in membrane potential. We discuss the possible role of the mot gene product as an ion gate controlled by a methylation-demethylation process in response to attractants and repellents acting through their chemoreceptors.