Background: No reliable variables to predict clinical or laboratory response to treatment in patients with rheumatoid arthritis were available until recently. Aim: To asses the potential predictive value of the Sharp s modified radiographic joint damage index for the assessment of clinical and laboratory response to a methylprednisolone i.v. pulse. Patients and Methods: Twenty-two patients suffering from rheumatoid arthritis received a single i.v. pulse of 1 g of methylprednisolone. Hand X-rays were taken at baseline and blindly scored by two trained radiologists. Clinical and laboratory variables were assessed at baseline and at weekly intervals up to 30 days plus a 60 days final evaluation. Improvement was defined as a 50% amelioration in 4 variables. Results: Assessment of radiographic scores had a high correlation between and within observers (intraclass correlation= 0.998). Sharp score did not reach statistical significance as global predictor for the inflammatory variable response to methylprednisolone. However, when the number of swollen joints was taken into account, patients with a low erosive score (Sharp "d 50) had a more prolonged clinical response, than patients with higher erosive score (Sharp >50) (Fisher test p= 0.023). It is of clinical importance to point out that among patients with high Sharp score there were also responders who reached a high level of improvement. A statistically significant correlation between the basal PCR serum titers and the radiographic score (p< 0.02) was observed. Conclusions: The number of swollen joints and other variables that consider joint structural changes should be considered for the assessment of rheumatoid arthritis patients