The significance of hepatic changes in methotrexate-treated RA patients is unclear at this time. In our group of RA patients, there was a slight increase in the incidence of triaditis and fat during methotrexate therapy. Disease duration greater than or equal to 10 years was associated with increased hepatic triaditis before treatment. Age greater than 50 years was associated with increased hepatic fat before and after treatment. It appears that patients' ages and duration of underlying RA account for some changes, independent of methotrexate therapy. Several of our patients changed from higher to lower histologic grade or had an apparent decrease in fibrosis, fat, or triaditis on the pathologists' reports and the blind readings of the repeat biopsies. This may be explained by sampling error. More importantly, some of these changes may not be of clinical significance. One report of methotrexate-induced cirrhosis in patients with psoriasis demonstrated that in all but one of 14 patients who continued receiving methotrexate the cirrhosis decrease or did not progress. This may also be true of the hepatic fibrosis seen in RA after methotrexate treatment. In this study, there did not appear to be changes seen on pretreatment liver biopsy that were predictive of subsequent fibrosis or cirrhosis. Our data indicate that pretreatment biopsy is unwarranted in a population similar to ours. However, our practice has been to try to avoid methotrexate in patients with diabetes, prior liver disease, alcoholism, or obesity because of previous reports suggesting that these patients are at increased risk for the development of cirrhosis. Only the above-mentioned patient, eventually diagnosed as having cirrhosis, might have been handled differently. Including the study, none of the approximately 700 RA patients in the literature having liver biopsies after methotrexate therapy have developed cirrhosis consequent to its use. Most of these had received a total dose of approximately 1,500 mg in small weekly doses, and alcohol was prohibited. Below this cumulative dose the risk of clinically significant liver damage in carefully selected patients is very low. In view of this experience, the recommendation that RA patients have liver biopsies after 1,500 mg of methotrexate (a holdover from the psoriasis literature) may be too conservative in low-risk RA patients, provided methotrexate is administered weekly and alcohol is prohibited. Recognizing that the absolute need for biopsy is unproven, a more realistic milestone for those choosing biopsy might be after each 2,000 to 2,500 mg.(ABSTRACT TRUNCATED AT 400 WORDS)