What rheumatologist has not been asked about the role of diet ifi the treatment of rheumatoid arthritis? Who among us has not responded that there is no evidence that diet treatment works? Editors are generally reluctant to publish negative studies, but the paper by Panush et al in this month's Arthritis and Rheumatism represents an honest attempt to address this problem. It will not, of course, satisfy the zealots who will complain that the initial fasting period was insufficient, or that proper dietary supplements were not included, or that high colonic enemas did not accompany the diet. Nevertheless, the question is important, particularly if we put aside our preconceptions and consider how dietary manipulation might prove beneficial. Is it possible that what we eat can influence the concentration of precursors of inflammatory mediators in cell membranes, enhance or diminish lymphocyte responses, or influence the concentration of endorphins or related substances in the blood stream or central nervous system? With these thoughts in mind, we have asked Moms ZifF to review the existing information about the effects of diet on immunologically related diseases of experimental animals, and to suggest how similar studies might be performed with patients who have rheumatoid arthritis.