Over the decades, nonsteroidal anti-inflammatory drugs (NSAIDs) and opioids are the most commonly used analgesics in the management of acute and chronic pain. In order to assess a possible antinociceptive interactions, the antinociceptive effects of rofecoxib p.o., a preferential inhibitor of cyclooxygenase-2, and tramadol-hydrochloride p.o., an atypical opioid analgesic, administered either separately or in combination, were determined using a rat model of arthritic pain. The data were interpreted using the surface of synergistic interaction (SSI) analysis and an isobolographic analysis to establish the nature of the interaction. The SSI was calculated from the total antinociceptive effect produced by the combination after subtraction of the antinociceptive effect produced by each individual drug. Female rats received orally rofecoxib alone (1.0, 1.8, 3.2, 5.6, 10.0, 17.8, 31.6 and 56.2 mg/kg), tramadol alone (1.8, 3.2, 5.6, 10.0, 17.8, 31.6 and 56.2 mg/kg) or 12 different combinations of rofecoxib plus tramadol. Five combinations exhibited various degrees of sub-additive (i.e. less than the sum of the effects produced by the each drug alone) antinociceptive effects (3.2 mg/kg tramadol with 7.8 mg/kg rofecoxib; 5.6 mg/kg tramadol with either 10.0 or 17.8 mg/kg rofecoxib; 10.0 mg/kg tramadol with either 10.0 or 17.8 mg/kg rofecoxib), whereas the other 7 combinations showed additive antinociceptive effects (i.e. the sum of the effects produced by each agent alone). Three combination of rofecoxib+tramadol (10.0+5.6, 10.0+10.0, and 17.8+5.6 mg/kg respectively) presented high sub-additive interactions (P<0.002: Q=9.5). The combination rofecoxib (17.8 mg/kg)+tramadol (10.0 mg/kg) caused gastric injuries less severe than those observed with indomethacin, but more severe than those obtained with rofecoxib or tramadol in single administration. The antinociceptive interaction of rofecoxib and tramadol suggests that combinations with these drugs may have no clinical utility in pain therapy.