Tetrabromobisphenol A (TBBPA) is an important flame retardant used primarily as a reactive chemical in the synthesis of printed circuit boards. In this application, it is incorporated into the epoxy and serves a vital role in fire safety. Despite its use primarily as a reactant with limited release to the environment, some concerns have been raised regarding exposure. The objective of this study was to develop non-cancer based toxicity values (reference doses, or RfDs) and to compare such to appropriate estimates of human exposure. An array of potential RfDs was developed using data reported as part of a recent 2-year bioassay conducted by the National Toxicology Program (NTP); these data indicated that chronic exposure to TBBPA induced hepatic and kidney lesions in male mice, and forestomach lesions in male and female mice. In rats, notable lesions included uterine hyperplasia and rete overii cysts at study termination, as well as decreased serum thyroxin levels after 13 weeks of exposure. These endpoints were modeled using benchmark dose (BMD) modeling; the 95% lower confidence interval values (BMDLs) were derived with a benchmark response set to 10% extra risk for dichotomous endpoints and one standard deviation for continuous endpoints. BMDL values were adjusted by allometric scaling and subsequently reduced by uncertainty factors to derive RfD values ranging from 0.03 to 3.0 mg/kg-day. In comparison, exposure estimates ranged from 0.0000003 to 0.00007 mg/kg-day based on an evaluation of exposures for infants (based on intakes due to breastmilk and soil/dust), and for children, adolescents, and adults (based on intakes due to diet, drinking water, and soil/dust ingestion). The total daily exposure estimates were generally driven by soil/dust ingestion for most receptors (and by breastmilk for infants) in reasonable maximum and upper-end regulatory default exposure scenarios. These data indicate that there is a sufficiently large margin of safety between potential RfD values and current exposure to TBBPA.