Metabolic models for U in adults recommended by Wrenn et al. (1985) and the International Commission on Radiological Protection (ICRP 1979a) were used to study the relationship between kidney burden and radiation dose from chronic ingestion of soluble 238U or natural U and whether current radiation standards for the public provide adequate protection against chemical toxicity from U in the kidney. We assumed that the threshold concentration for chemical toxicity is 1 microgram of U g-1 of kidney and that a safety factor of 10 should be applied in limiting kidney burdens for maximally exposed individuals in the general public. We found that a limit on annual effective dose equivalent of 1 mSv (0.1 rem) for chronic exposures of the public from all sources, as recommended by the ICRP (1985) and the National Council on Radiation Protection and Measurements (NCRP 1987), corresponds to concentrations of U in the kidney from chronic ingestion that exceed the assumed threshold for chemical toxicity of 1 microgram g-1 only for 238U using the metabolic model of the ICRP (1979a). However, using either metabolic model (ICRP 1979a; Wrenn et al. 1985), the predicted concentrations of U in the kidney exceeded the limit of 0.1 microgram g-1, based on the assumed safety factor for protection of the public, for both 238U and natural U. From these results, we concluded that chemical toxicity should be considered in developing health protection standards for the public for ingestion of soluble 238U or natural U. Environmental radiation standards for certain practices established by the U.S. Environmental Protection Agency and Nuclear Regulatory Commission (EPA 1987a, 1987b, 1987c, 1987d; NRC 1988a) are consistent with a limit on annual effective dose equivalent of 0.25 mSv (25 mrem) per practice. If the metabolic model of Wrenn et al. (1985) is assumed to be appropriate for chronic ingestion of soluble U in the environment, then the dose limit of 0.25 mSv corresponds to a concentration of 238U or natural U in the kidney that is below the assumed limit of 0.1 microgram g-1 for members of the public. Inhalation of soluble and insoluble U and ingestion of insoluble U were considered. Except for inhalation of soluble U, these modes of intake reduced predicted concentrations in the kidney per unit effective dose equivalent compared with values for ingestion of soluble U. Unresolved issues of importance for determining the significance of chemical toxicity relative to radiation dose in establishing limits on public exposures for U also are discussed.