The relationship between breast cancer and two trace elements, zinc (Zn) and copper (Cu), was investigated by means of an hospital based case-control study at Milan (Italy) and Montpellier (France). Variables concerning dietary intake of Zn and Cu (in Milan) and their blood levels (both in Milan and Montpellier) were measured. Dietary intake, evaluated through a questionnaire of the dietary history type, and blood levels of Zn and Cu were measured in 261 cases and 261 controls. Cu blood level showed a contradictory tendency in the two samples (higher in controls in Milan, higher in cases in Montpellier), which tended to lessen after adjustment for related variables. No odds ratios (OR) in the different quantiles, nor X2 for trend reached statistical significance. A sharp difference was evidenced on the opposite with regard to Zn blood values in cases and controls. In both samples Zn mean values are significantly higher in cases than in controls, and the difference remains significant in the two samples even after adjustment for related variables. The pooled OR computed from the two samples, after adjustment for known risk factors and related variables, reaches in the fourth quartile a value of 9.5 (CI: 4.9-18.2). Dietary intake of the two minerals (measured only in Milan sample) showed no difference between cases and controls, but a stronger relationship between dietary and blood Zn was evidenced in cases with respect to controls. The authors suggest that the higher Zn level in cases might be related with an higher incorporation of Zn in cancer cases and that the same mineral might play a possible role in tumor growth promotion.