In order to study central neuronal components involved in subcutaneous (s.c.) bee venom-induced persistent pain (a new tonic pain model), we use Fos immunostaining technique to study the spatial and temporal patterns of neuronal activity in the spinal cord of anesthetized rats. Following intraplantar bee venom injection, Fos-like immunoreactive (ir) neurons were only seen from L1 to S3 rostrocaudally with distinct distribution at L4-5 segments. At segments of L1-2 and S1-3, Fos-ir labelings were diffusely and symmetrically distributed on both sides of the gray matter; however, at L4-5 segments, Fos-ir neurons were densely localized in medial portion of laminae I-II, less densely in laminae V-VI and a few in laminae VII and X ipsilateral to the injection side. No Fos labeling was seen in ventral horn of the spinal cord at L4-5 segments. Fos protein began to express only within lamina I at 0.5 h, but increased over the whole dorsal horn at 1 h and reached peak labeling at 2 h after bee venom. Expression of c-Fos in laminae I-II decreased at 4 h, and completely disappeared at 24 h, however, labeling in laminae V-VI disappeared much slowly and existed even at 96 h after bee venom. Within laminae III-IV, Fos-ir neurons could not be seen at 0.5 h, but began to be seen at 1 h and appeared to exist even at 24 h after bee venom. Systemic morphine suppressed c-Fos expression dose-dependently in both superficial and deep layers of dorsal horn and the latter region was much more sensitive to morphine than the former one. The present results demonstrated that prolonged neuronal activities in superficial and deep layers of dorsal horn were essential to mediation of bee venom induced tonic pain and may have different roles in generation and/or modulation of spontaneous pain and hyperalgesia and allodynia.