Tumor necrosis factor-alpha (TNF-α) is clearly implicated in chronic pain, but its role for acute pain is not well characterized. This study evaluated changes in TNF-α gene expression associated with the onset of clinical pain, the development of inflammation, and the administration of prototypic analgesics in a well-characterized clinical model of pain and inflammation. A total of 232 oral mucosal biopsies were taken from 116 healthy volunteers who required the surgical extraction of impacted third molars. Subjects randomly received orally either rofecoxib (50 mg, QD), acetaminophen (1000 mg, QID), or placebo 60 min before surgery or intravenously received ketorolac (30 mg) 30 min before surgery. Total RNA extracted from each biopsy was used to analyze gene expression following acute inflammation and drug treatments using qRT-PCR (ABI Prism 7900 HT). TNF-α gene expression was significantly up-regulated (6-fold) in the tissues collected from the placebo group at 3 hrs post-surgery, the time of onset of moderate to severe pain. TNF-α gene expression was also increased in the 3 hrs post-surgery tissues in the ketorolac (3-fold), rofecoxib (3-fold) and acetaminophen (4-fold) groups. The increase was also observed in the 24 hrs drug-treated post-surgery tissues (2-3 folds) but with no significant differences across groups. These data demonstrate that TNF-α gene expression is up-regulated following tissue injury at time points associated with acute pain and the development of inflammation but is not significantly affected by prototype analgesic drugs. The up-regulation of TNF-α in the 24 hrs post-surgery tissues suggests continued participation in the inflammatory cascade. The failure of standard anti-inflammatory analgesics to suppress TNF-α and the wide inter-individual variation seen may explain, in part, the ceiling effect associated with drugs that act though suppression of COX pathways - in the face of the multiple mediators that drive the inflammatory cascade.