The four arsenic-containing chemical warfare agents (CWA) Adamsite (technical, 10-chloro-9-10-dihydrophenarsazine), Clark 1 (Diphenyl arsine chloride), Clark 2 (Diphenyl arsine cyanide), and Lewisite (2-chloro-ethenyl dichloro arsine) were evaluated toxicologically using an in vitro and an in vivo model. The CWA were tested in vitro, using human leucocytes, in order to evaluate their effects on cell proliferation and cell cycle kinetics assayed by flow cytometry. The concentration for total inhibition of cell proliferation for the CWAs ranged from 0.3 micrograms/mL (Lewisite) to 15 micrograms/mL (Clark 1). The results showed no differences in cell proliferation between the different stages of the S-phase either at no-effect or at effect concentrations when samples treated with the CWAs were compared with untreated controls. As2O5 was used as a reference, and the arsenic concentration required for total inhibition of the cell proliferation was 1.7 mg/mL. A lower arsenic concentration, 0.83 micrograms/mL, showed a decreasing proliferation ration in the different parts of the S-phase with Searly and Slate having the highest and lowest ratios, respectively. In addition, there was a positive correlation between the unlabeled S-phase cell ratio and the arsenic concentration, indicating cessation of DNA-synthesis. Conclusively, the examined CWAs exerts a toxicity more potent than arsenic, with respect to cell proliferation. This might indicate that the toxicity of these arsenic-containing CWAs only to a minor extent can be explained by their arsenic content. Because of their relatively good water solubility and ability to easily degrade to arsenic acid, organo-arsenics, e.g., Lewisite, pose less of an environmental threat than organo-arsenics that are sparingly soluble and chemically persistent, like Adamsite. The four CWAs were also tested in vivo in a dietary exposure study, using juvenile three-spined stickleback (Gasterosteus aculeatus L.) as test organism, with measurement of EROD-activity and studies of liver morphology. The results from the in vivo study indicated small effects, suggesting that a 10-week period of dietary exposure at tested dose levels (1 and 100 ng/ml) affects juvenile three-spined stickleback only to a minor extent.