No comprehensive data on sources or risk factors of cadmium exposure in Ukrainian children are available. In this we measured the blood levels of cadmium among 80 Ukrainian children and evaluated sources of exposure. A nested case-control study from a prospective cohort of Ukrainian 3-year-old children was conducted. We evaluated predictors of elevated blood cadmium using a multivariable logistic regression model. The model included socioeconomic data, parent occupation, environmental tobacco smoke, hygiene, body-mass index, and diet. Dietary habits were evaluated using the 1992 Block-NCI-HHHQ Dietary Food Frequency survey. Elevated cadmium was defined as blood levels in the upper quartile (0.25 microg/L). The mean age for all 80 children was 36.6 months. Geometric mean cadmium level was 0.21 microg/L (range = 0.11-0.42 microg/L; SD = 0.05). Blood cadmium levels were higher among children taking zinc supplements (0.25 vs 0.21 microg/L; P = 0.032), children who ate sausage more than once per week (0.23 vs 0.20; P = 0.007) and children whose fathers worked in a by-product coking industry (0.25 vs 0.21; P = 0.056). In the multivariable model, predictors of elevated blood cadmium levels included zinc supplementation (adjusted OR = 14.16; P < 0.01), father working in a by-product coking industry (adjusted OR = 8.50; P = 0.03), and low body mass index (<14.5; adjusted OR = 5.67; P = 0.03). This is the first study to indicate a strong association between elevated blood cadmium levels and zinc supplementation in young children. Whole-blood cadmium levels observed in this group of Ukrainian children appear to be similar to those reported in other Eastern European countries.