Sensitivity of the assay for Cu,Zn superoxide dismutase 3 (SOD3), the predominant form of SOD in serum, can be increased, and interferences caused by low-molecular-weight substances in the serum can be reduced by conducting the assay at pH 10 with xanthine/xanthine oxidase and acetylated cytochrome c (cyt c) as superoxide generator and detector, respectively. Serum SOD3 activity was assayed under these conditions in an experiment where weanling, male rats were fed diets for 6 weeks containing 3, 5 and 15 mg Zn/kg with dietary Cu set at 0.3, 1.5 and 5 mg Cu/kg at each level of dietary Zn. Serum SOD3 responded to changes in dietary Cu but not to changes in dietary Zn. A second experiment compared serum SOD3 activity to traditional indices of Cu status in weanling, male and female rats after they were fed diets containing, nominally, 0, 1, 1.5, 2, 2.5, 3 and 6 mg Cu/kg for 6 weeks. Serum SOD3 activity was significantly lower (P < .05) in male rats fed diets containing 0 and 1 mg Cu/kg and female rats fed diet containing 0 mg Cu/kg compared with rats fed diet containing 6 mg Cu/kg. These changes were similar to changes in liver Cu concentrations, liver cyt c oxidase (CCO) activity and plasma ceruloplasmin in males and females. Serum SOD3 activity was also strongly, positively correlated with liver Cu concentrations over the entire range of dietary Cu concentrations (R(2) = .942 in males, R(2) = .884 in females, P < .0001). Plots of serum SOD3 activity, liver Cu concentration, liver CCO activity and ceruloplasmin as functions of kidney Cu concentration all had two linear segments that intersected at similar kidney Cu concentrations (18-22 microg/g dry kidney in males, 15-17 microg/g dry kidney in females). These findings indicate that serum SOD3 activity is a sensitive index of Cu status.