Stress exposure resulted in brain induction of immediate-early genes (IEGs), considered as markers of neuronal activation. Upon repeated exposure to the same stressor, reduction of IEG response (adaptation) has been often observed, but there are important discrepancies in literature that may be in part related to the particular IEG and methodology used. We studied the differential pattern of adaptation of the IEGs c-fos and arc (activity-regulated cytoskeleton-associated protein) after repeated exposure to a severe stressor: immobilization on wooden boards (IMO). Rats repeatedly exposed to IMO showed reduced c-fos mRNA levels in response to acute IMO in most brain areas studied: the medial prefrontal cortex (mPFC), lateral septum (LS), medial amygdala (MeA), paraventricular nucleus of the hypothalamus (PVN) and locus coeruleus. In contrast, the number of neurons showing Fos-like immunoreactivity was only reduced in the MeA and the various subregions of the PVN. IMO-induced increases in arc gene expression were restricted to telencephalic regions and reduced by repeated IMO only in the mPFC. Double-labelling in the LS of IMO-exposed rats revealed that arc was expressed in only one-third of Fos+ neurons, suggesting two populations of Fos+ neurons. These data suggest that c-fos mRNA levels are more affected by repeated IMO than corresponding protein, and that arc gene expression does not reflect adaptation in most brain regions, which may be related to its constitutive expression. Therefore, the choice of a particular IEG and the method of measurement are important for proper interpretation of the impact of chronic repeated stress on brain activation.