Conference calls are handled using distributed processing. In one embodiment, the invention includes receiving sets of telephone audio signals from conference subnodes, the telephone audio signals corresponding to subscribers of the conference call, selecting an output set of telephone audio signals from all of the received telephone audio signals, and transmitting the output set to the conference subnodes. In another embodiment, the invention includes receiving a set of telephone audio signals, each signal being received from a subscriber node of a single conference call, selecting a subset of the set of signals, transmitting the selected subset of signals to a conference node, receiving a second set of telephone audio signals from the conference node, and transmitting the second set to the subscriber nodes.