Purpose: Cd is considered as a genotoxic carcinogen for which a threshold can be identified. This threshold has, however, not been established and the shape of the relationship between Cd exposure and genotoxic effects is unknown. The aim of the present study was to analyse the shape of the dose-response relationship for the genotoxic effects of Cd in occupational settings. Methods: The study has a cross-sectional design and includes 60 healthy male and female workers with known Cd exposure selected from two plants manufacturing or recycling nickel-Cd batteries. The frequency of MN was measured in circulating lymphocytes, and related to internal Cd doses (Cd-B, Cd-U). Determinants of MN frequency were traced by multivariate regression analysis. Results: Cd exposure covered a wide range as measured by Cd-B (0.02-1.26 mu g/dL), Cd-U (0.26-15.80 mu g/g creat) and seniority in the plant (1-42 years). Gender was the only parameter significantly associated with MN frequency, women having on average 8.5 additional MN/1000 BN cells compared to men. Cd-B, Cd-U or Ni-U did not influence MN frequency when adjusted for gender and other potential confounders. Conclusion: This finding is consistent with the existing knowledge on the mechanisms governing the genotoxic activity of Cd, which are all non-stochastic and thresholded. The threshold for systemic genotoxic effects of Cd is thus beyond the range of internal exposure considered in the present investigation.