Abstract The capacity of adult mammalian gustatory and olfactory receptor cells to regenerate and make synaptic reconnections provides examples that may be useful in initiating replacement of other kinds of sensory receptor cells. The sensory code for taste quality may not be degraded by taste receptor cell turnover because axons probably recouple to the appropriate type of new receptor cell by axon-receptor cell affinity. Experiments on the development and regeneration of taste receptor cells suggest that they regenerate and turn over by recapitulating the late but not the early steps in taste bud development. To evaluate the replacement of vertebrate olfactory receptors, we began by characterizing the spatial pattern of primary olfactory projections in rainbow trout. contiguous clusters of HRP-labeled olfactory receptor neurons (ORN) make highly divergent projections to the olfactory bulb. Retrograde transport of fluorescent latex beads revealed that a given restricted site in the glomerular layer received axons from ORNs widely scattered in the epithelium. Hence, ORN axons do not form point-to-point or regional topographic maps. Rather, the olfactory epithelial sheet makes a plane-to-point or holographic-like projection, since any given point in the glomerular layer receives information from the entire olfactory epithelial plane. Receptor cells that reacted with the lectin pokeweed agglutinin were highly dispersed in the olfactory epithelium with axons widely scattered in the olfactory nerve. Yet, as a consequence of the extensive reaggregation of axons at the nerve-bulb interface, the lectin-positive axons fasciculated and converged into a subregion of the glomerular layer. This is consistent with the view that there are many dispersed sets of ORNS whose axons fasciculate and converge to subregions of the olfactory bulb.