This is a critical overview of existing scoring methods including their basis, their measurement systems, their advantages and disadvantages. Conventional radiography is still-since available everywhere, interpretable and cost effective-the best imaging method to evaluate the course of RA. Scoring methods are designed to semiquantitatively measure radiographically visible changes, especially erosive destruction and-in part-cartilage loss. These methods are well validated, reproducible, and yield similar results in clinical trials. Most methods overestimate early changes and have a clear ceiling effect. Within the time frame of clinical trials radiographic evaluation is not very sensitive to change since the progression of destruction in RA is relatively slow. Moreover, small erosions cannot be detected if they are not at the margin of the bone or if they are superimposed by other bones. Within the Larsen method the definition of grade 1 by soft tissue swelling is disadvantageous: soft tissue swelling is difficult to identify on X-rays, it is a measure of disease activity and not of destruction and is quickly reversible. Joint space narrowing, measured with Sharp's method and its modifications, may be caused by misprojection due to soft tissue swelling with flexion or subluxation of the joint rather than cartilage loss. Since the measurement error of a scoring method is very much dependent on the severity of the disease of the patient population and on the quality of the radiographs, the measurement error and thereby the minimal detectable change (MDC) should be stated for every single clinical trial. Conventional radiographs and scoring methods are still indispensable to measure the influence of treatment on radiographic progression in RA. A future task will be to include in scoring methods the rating of reparative changes which cannot be scored so far.