A brief history of the development of scoring methods to represent the extent of radiographic abnormalities in rheumatoid arthritis shows that grading systems have become more detailed with time. Taken together with the observation that physicians continue to modify scoring methods it is clear that we have not yet arrived at the ideal method. Issues to be considered in evaluating scoring methods include which abnormalities and which joints to score and what scale to use. Data from one early trial indicate that erosion and joint space narrowing scores do not progress in lockstep and progression in one does not predict progression in the other; this leads to the conclusion that including both features in a scoring method is important. Data from the same trial raise the possibility that joint space narrowing scores for the proximal interphalangeal finger joints may not be helpful in detecting treatment differences, illustrating that further investigation of which joints to include in scoring is needed. The possibility that an expanded scoring scale might be more sensitive to recording a change in radiographic damage also was raised. A simple method of determining which patients show progression of damage was proposed based on assessing the distribution of negative progression scores (baseline erosion scores subtracted from followup scores adjusted for time). It was pointed out that radiographic assessment has established that a few drugs slow the progression of joint damage. Radiographic evaluation of disease progression should continue to be utilized as an outcome measure in evaluating new therapies.