The velocity of 180° domain walls in orthorhombic BaTiO3has been measured over a range of applied fields below 1 kV/cm in dc and in pulsed field conditions. From these measurements it was seen that the domain wall behavior was qualitatively very similar to that reported by Miller and Savage for the tetragonal case. The main difference was the more pronounced effect of the surface layer on the velocities of the walls in the orthorhombic crystals. Accordingly, a number of additional experiments were performed which gave more information on the nature of this surface layer. As a result a completely new model of the surface layer is proposed in which the activation field &dgr;Lin the layer is much higher than the activation field &dgr;Bin the bulk of the crystal. The predictions from this model are in satisfactory agreement both with the results reported here for orthorhombic BaTiO3and with all the evidence available from other sources of the behavior of tetragonal BaTiO3. Comparison is made with previously accepted models and it is shown that these fail to explain many of the experimental results both of the orthorhombic and of the tetragonal cases.