OBJECTIVES: To identify predictors of tibial and dimercaptosuccinic acid (DMSA) chelatable lead in 543 organolead manufacturing workers with past exposure to organic and inorganic lead. METHODS: In this cross sectional study, tibial lead (by 109Cd K-shell x ray fluorescence), DMSA chelatable lead (4 hour urinary lead excretion after oral administration of 10 mg/kg), and several exposure measures were obtained on study participants, mean (SD) age 57.6 (7.6) years. RESULTS: Tibial lead concentrations ranged from -1.6 to 52.0 micrograms lead/g bone mineral, with a mean (SD) of 14.4 (9.3) micrograms/g. DMSA chelatable lead ranged from 1.2 to 136 micrograms, with a mean (SD) of 19.3 (17.2) micrograms. In a multiple linear regression model of tibial lead, age (p < 0.01), duration of exposure (p < 0.01), current (p < 0.01) and past (p = 0.05) cigarette smoking, and diabetes (p = 0.01) were all independent positive predictors, whereas height (p = 0.03), and exercise inducing sweating (p = 0.04) were both negative predictors. The final regression model accounted for 31% of the variance in tibial lead concentrations; 27% was explained by age and duration of exposure alone. DMSA chelatable lead was directly associated with tibial lead (p = 0.01), cumulative exposure to inorganic lead (y.microgram/m3, p = 0.01), current smoking (p < 0.01), and weight (p < 0.01), and negatively associated with diabetes (p = 0.02). The final model accounted for 11% of the variance in chelatable lead. When blood lead was added to this model of DMSA chelatable lead, tibial lead, cumulative exposure to inorganic lead, and diabetes were no longer significant; blood lead accounted for the largest proportion of variance (p < 0.001); and the total model r2 increased to 19%. CONCLUSIONS: The low proportions of variance explained in models of both tibial and chelatable lead suggest that other factors are involved in the deposition of lead in bone and soft tissue. In epidemiological studies of the health effects of lead, evaluation of associations with both these measures may allow inferences to be made about whether health effects are likely to be recent, and thus potentially reversible, or chronic, and thus possibly irreversible. The data also provide direct evidence that in men the total amount of lead in the body that is bioavailable declines with age.