Tris(2,3-dibromopropyl)phosphate (Tris-BP) was found to be metabolized by liver microsomes obtained from untreated and phenobarbital-pretreated rats. Metabolites of Tris-BP, whose formation was dependent on NADPH and oxygen, included bromide ion and bis(2,3-dibromopropyl)phosphate (Bis-BP). The rates of formation of these metabolites were markedly increased in liver microsomes isolated from phenobarbital-pretreated rats compared to microsomes from untreated rats. In the presence of either SKF 525-A or metyrapone, the formation rates of bromide ion and Bis-BP were decreased, whereas alpha-naphthoflavone had no effect. The effects of the various treatments on bromide release and Bis-BP formation paralleled those that have been previously observed with respect to the activation of Tris-BP to mutagenic and covalently protein bound metabolites. Furthermore, rates of oxidative debromination of several Tris-BP analogs directly correlated with their respective mutagenicities. Addition of glutathione (GSH) to microsomal incubations of Tris-BP increased bromide release substantially over control, values but had no effect on Bis-BP formation. On the other hand, the addition of GSH to microsomes decreased covalent binding and mutagenicity of Tris-BP with increased formation of water soluble metabolites. GC/MS analysis of ethyl acetate extracts from incubations of rat liver microsomes with Tris-BP identified 2-bromoacrolein (2-BA) as a metabolite. Introducing deuterium at the carbon atom number 1 of the propyl moiety of Tris-BP had no effect on either bromide release or mutagenicity, whereas the analog labelled at carbon atom 3 showed significant isotope effects on both activities. In contrast, deuterium substitution at carbon atom 2 gave a significant isotope effect on bromide release, but not on mutagenicity. The data indicate that Tris-BP can be metabolized by rat liver microsomes to Bis-BP and 2-bromoacrolein catalyzed by cytochrome P-450 in a process liberating bromide ions. Further, the results are consistent with oxidation at the terminal carbon atom of Tris-BP thereby forming 2-bromoacrolein, which is postulated to be the metabolite mainly responsible for Tris-BP mutagenicity.