The coding system described minimises the d.c. component in the transmitted binary signal by inverting characters when this will reduce the accumulated mark and space disparity. An added digit indicates when inversion has taken place. Conversion to and from unrestricted binary coding is relatively simple.