Hemoglobin adduct measurements provide a useful tool for the development of models for risk assessment of chemicals. The measurements can be used to identify reactive metabolites and to determine the in vivo doses of these metabolites. Thus, the need to use default assumptions with regard to animal metabolism and dose-response relationships can be reduced. The research presented here involved the study of a number of important industrial chemicals, primarily 1,3-butadiene, glycidyl ethers, acrylonitrile and acrylamide. Many of these chemicals were shown to be mutagenic in several tests (in vivo/in vitro) and are carcinogenic in experimental animals. The compounds or their metabolites have the ability to react with DNA to form adducts. Formation of such adducts is believed to be responsible for the development of mutation and cancer. The main aim has been to study relationships between exposure and in vivo dose. In some cases analytical methods to measure hemoglobin adducts were already available (for acrylonitrile and acrylamide) or were relatively straightforward (for some glycidyl ethers). In other cases considerable method development was required (for reactive metabolites of 1,3-butadiene, allyl glycidyl ether and acrylamide). The hemoglobin adducts analyzed in the present study are formed from the electrophilic moiety of the compound of interest with N-terminal valine in hemoglobin. The analysis and quantification of these adducts was achieved by derivatization with pentafluorophenyl isothiocyanate, in some cases followed by acetylation, and analysis of the corresponding pentafluorophenylthiohydantoin derivatives by GC/MS or GC/MS-MS. The methods have been applied for dosimetry in experimental animals as well as for biomonitoring exposure in human populations. Discussions in the literature on the carcinogenicity of 1,3-butadiene have been focused on its metabolites diepoxybutane and 1,2-epoxy-3-butene. Our results show that 3,4-epoxy-1,2-butanediol is an important metabolite of 1,3-butadiene. Another important outcome of the studies was that the double bond of allyl glycidyl ether is metabolized to a reactive epoxide in mice. This means that a bifunctional alkylating agent, diglycidyl ether, could possibly be formed. Bifunctional alkylating agents are usually more biologically effective than monofunctional compounds. Data obtained from this research will help to better understand the mechanism of toxicity of the chemicals under study and to provide an improved basis for risk estimation.