Abstract: A method of sampling marks on a transmission line and adjusting their peak amplitude to compensate for attenuation of signal strength on the line and of "voltage addition" caused by multiple terminals transmitting on the line. For networks meeting the CCITT I-Series Recommendation, two B-channels and a D-channel are transmitted in frames over the line; the B-channel forming the best basis for threshold adaptation, the D-channel the next best basis and the F-bit preceeding each frame used in the absence of either B- or D-channel signals. Accordingly, a B-bit mark sample is taken if possible, otherwise a D-bit mark sample; or if neither is present an F-bit mark sample; the adaptive threshold preferably being set at 55% of the peak value of the sample. The method is readily implemented as a "state machine" and consequently can be constructed from a programmable logic array.