Zinc and copper have important effects on T cell mediated immunity and on neutrophil function, but it is not known how the causes or effects, of low serum zinc and high serum copper relate to the clinical picture of rheumatoid arthritis (RA). In this study serum zinc and copper determined by flame atomic absorption spectrometry and 30 other clinical, immunological, and laboratory parameters in 60 patients with RA were analysed by stepwise multiple linear regression analysis. Joint score index, rheumatoid factor titre, seropositivity, haemoglobin, and C reactive protein (CRP) were among the nine independent variables which together predicted 73% of the serum zinc variation. This suggests that there is an association between the immune-inflammatory rheumatoid process and the serum zinc concentration. CRP alone had only a 3% independent predicting value for serum zinc, however. This suggests that metallothionein mediated sequestration in the liver, induced by interleukin 1, is not an important explanatory factor in a cross sectional study of chronic inflammation. Furthermore, serum zinc did not have any predictive value at all for serum copper concentration. This does not support the hypothesis suggesting that serum zinc deficiency leads to high serum copper by inducing gastrointestinal metallothionein and high caeruloplasmin.