The olfactory epithelium and its neuronal population are known to have a substantial capacity to recover after either direct injury or damage to the olfactory nerve. However, the mechanisms underlying that capacity for recovery, and indeed the limits on the recovery process, are not well understood. The aim of this study is to describe in detail the way in which the olfactory epithelium reconstitutes after direct injury. Adult male rats were exposed to 330 ppm methyl bromide (MeBr) gas for a single 6—hour period. The exposure destroys all of the neurons and sustentacular cells in over 95% of the olfactory epithelium of food‐restricted rats and in over 90% of the epithelium in ad‐libitum‐fed rats of the same weight, yet substantial recovery of the olfactory epithelium occurs. In response to the lesion, cellular proliferation increases markedly beginning between 24 and 48 hours, peaks at 1 week, and persists at levels higher than the control level for more than 4 weeks after MeBr exposure. Even though proliferation accelerates promptly, the beginning of neuronal reconstitution is delayed; only few immature neurons are observed 3 days after the lesion, yet they reappear in large numbers by the end of the first week. The first mature neurons emerge between 7 and 14 days after lesion and increase to near normal numbers by 4–6 weeks. In association with the restoration of neuronal population, basal cell proliferation returns to control levels between 4 and 6 weeks after damage. Likewise, sustentacular cells, identifiable by anticytokeratin 18 labeling, reappear rapidly and reform a distinct lamina in the superficial aspect of the epithelium. They closely resemble their counterparts in control epithelium with regard to disposition and shape by 3 weeks after lesion and with regard to expression of olfactory‐specific cytochrome P450s by 8 weeks. Thus, most areas of the epithelium are restored to a near normal appearance and cellular composition by the end of 8 weeks, suggesting that the MeBr paradigm for lesioning epithelium offers significant advantages over techniques such as Triton X‐100 or ZnSO₄ irrigation. However, not all measures of epithelial status are normal even at 8 weeks. Immature neurons remain slightly more numerous than normal at this time. Furthermore, some areas of the olfactory epithelium do not recover after MeBr lesion and are replaced by respiratory epithelium. By documenting the events and the extent of olfactory epithelial reconstitution after direct injury, the present results serve as a foundation for future studies designed to identify the mechanisms underlying epithelial recovery and the difference between areas of epithelium that reconstitute fully after lesion and those where the recovery process fails. © 1995 Wiley‐Liss, Inc.