Twelve patients with either rheumatoid or psoriatic arthritis were injected with a 10-mg bolus dose of methotrexate (MTX) either intramuscularly (n = 6) or intravenously (n = 6) and the MTX concentration in their sera was determined by radioimmunoassay. MTX concentration--time data fitted triexponential equations. Doses injected intramuscularly were rapidly and completely absorbed. There were no significant intergroup differences in drug mean t 1/2, volume of distribution, and total body clearance. In nine patients serum MTX concentrations remained above the suggested critical level of 0.01 mumol throughout the 24-hr study irrespective of the route of administration, but MTX did not cumulate in the serum. We conclude that the behavior of low doses of MTX in patients with arthritis closely resembles the pattern in patients receiving intermediate and high doses for the treatment of neoplasia.