We studied the grooming response to lateral ventricle injection of CRF in both rats and mice under similar conditions. One microgram of CRF ICV induced a pronounced increase (3- to 4-fold) in the frequency of self-grooming in rats, but only a much smaller (less than 20%) increase in mice. The minimum effective dose of CRF in rats was 300 ng. Although ACTH1-24 induced less grooming in mice than in rats, the difference in potency did not appear to be sufficient to explain the differences between the effectiveness of CRF in the two species. Whereas ACTH increased all types of grooming scored. CRF increased all forms of grooming except flank scratching with the hind limb. The major effect of CRF was to increase the number of episodes of grooming, whereas ACTH1-24 tended to prolong the length of individual episodes. The excessive grooming induced by ICV CRF was not affected by prior treatment with dexamethasone, suggesting that the increased grooming was not due to secondary release of ACTH from the pituitary. Nevertheless, ICV CRF might induce grooming by releasing MSH/ACTH from cerebral storage sites. CRF-induced grooming, like ACTH-induced grooming, was inhibited by naloxone pretreatment. Despite the small qualitative differences, CRF-induced grooming could be due to secondary release of ACTH.