A case-control study to investigate whether women employed in nickel-exposed work areas in early pregnancy are at elevated risk of spontaneous abortion (SA). Data about pregnancy outcome and maternal factors were obtained about each delivery and SA from women in selected work places. Each pregnancy record was assigned a categorical nickel (Ni) exposure rating according to the women's occupations at pregnancy onset. The guidelines were the water-soluble Ni subfraction of the inhalable aerosol fraction obtained by personal monitoring for nickel- and copper-refinery workers or/and measured urinary-Ni concentrations. The unadjusted odds ratio for the association between the maternal exposure to Ni and an SA for Ni-exposed women was 1.38 (95% confidence interval: 1.04-1.84), and the adjusted was 1.14 (0.95-1.37). In conclusion, there was no statistical association between maternal occupational exposure to water-soluble Ni in early pregnancy and the risk of self-reported SA. The findings do not exclude the possibility of a weak excess risk, or a risk in the first weeks of pregnancy.