Evidence of the analgesic effects of nonsteroidal antiinflammatory drugs in human diseases is easy to collect. However, demonstration of differences in the activity of NSAIDs according to the route of administration is considerably more difficult. Forty patients with rheumatoid arthritis were given either one intramuscular injection of 100 mg ketoprofen with two placebo capsules or one intramuscular injection of placebo with two 50 mg ketoprofen capsules. Analysis of changes in pain severity over the six-hour study period failed to disclose any statistically significant differences between the two groups. However, time to peak effectiveness (Tmax) was significantly shorter with the intramuscular route (194 +/- 118 minutes) than with the oral route (276 +/- 111 minutes) (p = 0.029). Although intramuscular NSAID therapy is rarely warranted in RA patients, these findings suggest that RA may serve as a clinical model for evaluating the effectiveness of a NSAID according to the route of administration.