Methotrexate (MTX) has proven to be efficacious in the treatment of rheumatoid (RA), but it remains to be proven whether it can slow disease progression, as determined radiographically, in comparison with other disease modifying antirheumatic drugs (DMARD). We performed a meta-analysis of the available data to answer this question. A literature search, including abstracts, was conducted and inclusion criteria developed (description of patients, accountability of patients, inclusion of a control group of patients, specified radiographic endpoint, and appropriate reading of the radiographs). Publications were scored on a scale of 0 to 5 with a score > or = 3 required for inclusion in the study. For abstracts selected, additional data were obtained directly from the investigators. Data for 353 MTX treated and 205 non-MTX-DMARD treated patients with RA were gathered. Not all publications used the same scoring system, so some assumptions were required to analyze the combined data. Only the erosion score was included since not all publications included a reading of the joint space. All scores were transformed into Sharp scores (Arthritis Rheum 1985;28:1449), including the important contributions of 3 Larsen scored publications. Finally a monthly rate of disease progression was computed. Several comparisons were made. Overall, the rates of disease progression were similar for MTX and non-MTX-DMARD treated patients with RA. The non-MTX-DMARD treated patients with RA were separated into a group treated with gold salts (oral or parenteral) and a group treated with azathioprine with each group compared to the MTX treated patients. MTX had slower rates of disease progression than azathioprine, (rates 0.004 vs 0.012) but not slower rates than gold salts (0.008 vs 0.008). Despite its efficacy, the possible role of MTX in slowing disease progression more than other DMARD, as determined radiographically, appears to be evident only when compared to azathioprine.