Although simple in concept, reliable antidromic identification of efferent populations poses numerous technical challenges and is subject to a host of sampling biases, most of which select against the detection of the neurons with slowly conducting axons. This problem is particularly acute in studies of the neocortex. Many neocortical efferent systems have large sub-populations with very slowly conducting, nonmyelinated axons and these elements have been relatively neglected in antidromic studies of neocortical neurons. The present review attempts to redress this problem by analyzing the steps that must necessarily precede antidromic identification and the sampling biases associated with each of these steps. These steps include (1) initial recognition that the microelectrode is near a neuron; (2) activation of the efferent axon via the stimulating electrode; (3) conduction of the antidromic impulse from stimulation site to soma; (4) detection of the antidromic spike in the extracellular record and (5) discriminating antidromic from synaptic activation. Experimental strategies are suggested for minimizing the sampling biases associated with each of these steps; most of which can be reduced or eliminated by appropriate experimental procedures. Careful attention to such procedures will make it possible to better understand the nature and function of the information flow along the very slowly conducting axonal systems of the neocortex.