[en] In order to examine the effect of estrogen on facial pain, we first compared the face-rubbing evoked by a formalin injection in the lip of aromatase-knockout (ArKO) mice, lacking endogenous estrogen production, 17 beta-estradiol-treated ArKO mice (ArKO-E2) and wild-type (WT) littermates. During the 'acute' phase of pain the time spent rubbing was similar in the three groups, whereas during the following 'interphase' and the second phase of pain, grooming was increased ill ArKO mice. Estradiol-treatment restored a behaviour similar to WT group. To better understand estrogens modulation on pain processes, we examined changes in 5-HT and CGRP innervations of trigeminal nucleus caudalis (TNC) in ArKO, ArKO-E2 and WT groups sacrified during the interphase. Whereas serotonin and CGRP immunoreactivities were comparable in WT and ArKO non-injected control groups, our data showed that 9 min after formalin injection, the density of serotoninergic terminals increased significantly in WT, but not in ArKO mice, while that of CGRP-immunoreactive fibers was lower in WT than in ArKO mice on the injected side. Estradiol-treatment only partially reversed these changes in ArKO-E2 mice. We conclude that estrogen deprivation in ArKO mice can be responsible for increased nociceptive response and that it is accompanied by transmitter changes favouring pro- over anti-nociceptive mechanisms in TNC during interphase of the formalin model. That estradiol-treatment completely reverses the behavioural abnormality Suggests that estrogens absence produces chiefly functional activation-dependent changes. However, the fact that the immunohistochemical abnormalities were not totally normalized by estradiol-treatment suggested that some permanent developmental alterations may occur in ArKO mice. (c) 2005 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved.