We have studied the formation of a resistive region in the capillary during DNA separation. This effect is caused by an unequal change in the mobilities of cations and anions at the interface between the running buffer solution and the capillary. We studied the motion of the resistive region boundary by sequential removal of portions of the affected capillary end. We found that in the process of developing the resistive region the distribution of the electric fields in the capillary changes from uniform to extremely nonuniform, with a very high field (above 1 MV/cm) in the resistive region and a reduced field (80 MV/cm) in the rest of the capillary. Though theoretically a resistive region may appear either at the anode (detection) or the cathode (injection) end of the capillary, all previous publications report the formation of the resistive region at the cathode side. In our experiments, however, the anomalous region is formed at the anode. Thus, the separated DNA peaks move towards the slowly progressing resistive region. Our results indicate that the DNA is stopped at the boundary and does not enter the region. When the resistive region is clipped off the peak motion resumes. This suggests that there exists a potential barrier at the resistive layer boundary that prevents the drift of the peaks towards the anode. The formation of the resistive region interferes with a normal separation process causing a gradual decrease of the capillary current and the deceleration and eventual quenching of the peak motion. For the ABI chemistry, we experimented with adding polymers to the electrode buffer to equate the transference numbers for anions and cations, and found the conditions at which this effect is completely eliminated.