Lipopolysaccharide (LPS)-induced hyperalgesia and the role of cyclooxygenase (COX) isoforms in acute and chronic nociceptive assays have been well established. However, the role of COX isoforms in LPS-induced hyperalgesia in the formalin test is not clear. Thus, the present study was undertaken to characterize the time course of formalin-induced nociceptive response in LPS-pretreated mice and to investigate possible effects of COX inhibitors to address the potential role of COX isoforms in LPS-induced hyperalgesia in the formalin test. All the animals showed typical biphasic response to formalin challenge. At 0 hr (immediately) and 4 hr after LPS pretreatment, animals did not show any alteration in formalin-induced tonic pain. However, 12 and 16 hr after LPS pretreatment, there was a significant increase in the late phase of formalin-induced nocifensive response as compared to control mice. Treatment with intravenously administered ketorolac (a nonselective COX inhibitor) significantly and dose-dependently inhibited the late phase of formalin-induced nociceptive behaviour in saline and LPS-pretreated mice. In contrast, parecoxib (prodrug of valdecoxib, a selective COX-2 inhibitor) or dexamethasone (COX-2 transcription inhibitor), when administered intravenously or intraperitoneally, respectively, did not show antinociceptive effect in the formalin test in saline-pretreated mice. However, both the agents significantly and dose-dependently decreased the late phase nociceptive behaviour of the formalin test in LPS-pretreated mice to the level of the animals that received saline pretreatment. These results suggest that induction of COX-2 by proinflammatory mediators and subsequent release of prostaglandins could be responsible for LPS enhancement of formalin-induced nocifensive behaviour and supports an important role of COX-2 in LPS-induced hyperalgesia in the formalin test.