The addition of the pore forming colicin A to Escherichia coli cells results in an efflux of cytoplasmic potassium. This efflux is preceded by a lag time which is related to the time needed for the translocation of the toxin through the envelope. Denaturing the colicin A with urea, before adding it to the cells, did not affect the properties of the pore but decreased the lag time. After renaturation, the lag time was similar to that of the native colicin. This suggests that the unfolding of colicin A accelerates its translocation. The addition of trypsin, which has access neither to the periplasmic space nor to the cytoplasmic membrane, resulted in an immediate arrest of the potassium efflux induced by colicins A and B. The possibility that trypsin may act on a bacterial component required for colicin reception and/or translocation was ruled out. It is thus likely that the arrest of the efflux corresponds to a closing of the pores. This long distance effect of trypsin suggests that part of the polypeptide chain of the colicins may still be in contact with the external medium even when the pore has formed in the inner membrane.