Over the last decade, evidence from experimental studies on neuronal migration in non-human primates has accumulated to the point where it can significantly amplify our understanding of the normal and pathological development of the human telencephalon. Systematic analysis of neuron genesis by the method of H3-thymidine autoradiography shows that in rhesus monkeys all neurons destined for the neocortex are generated near the surface of the lateral ventricle during a two-month period in the middle of gestation. Following their last cell division, young neurons migrate outwards across the cerebral wall to the developing cortical mantle, a journey that requires one to three days at early stages of neurogenesis, or more than two weeks towards the end of cortical development. From the very beginning, the basic columnar organization of the neuroepithelium favours radial migration. During later stages, when the primate telencephalic wall expands unevenly in thickness and surface area and begins to form primary fissures and cerebral promontoria, young neurons migrate to their cortical destinations in apposition to fascicles of radial glial fibres which span the full distance between the ventricular and pial surface. Furthermore, it appears that several generations of neurons all originate in the same restricted location at the ventricular surface, migrate along the same glial fascicles and consequently accumulate in the same radial cortical 'columns' in which, as a rule, somas of later generated neurons take positions external to the somas of their predecessors. It is proposed that fascicles of radial fibres (a) facilitate neuronal migration to the distant cortical plate through a complex assembly of closely-packed cells and processes that compose the developing primate telencephalon; (b) provide constraints which preserve a radial alignment of clonally related neurons in cortical columns; and (c) reproduce the mosaicism of the germinal ventricular zone at the expanded and curved cerebral surface.