Abstract Measurements of the membrane potentials of keratin membranes have been made using KCl and NaCl and the results show how the selectivity of the membranes can be altered by soaking in suitable buffer solutions. The results support Toerell's theory of membrane behavior in neutral salts, and some membrane potential measurements using HCl solutions are compared with Gilbert and Rideal's predictions. A qualitative theory is suggested for the mobility-pH curves obtained by electrophoresis. It also explains why the isoelectric point of keratin is not at pH 7, the point of zero membrane selectivity.