The water insoluble residues of some carotenoid-rich fruits and vegetables, such as apricots, oranges, brussels sprouts, carrots, yellow-red peppers, and tomatoes, were sequentially extracted with n-hexane, dichloromethane, acetone, and 2-propanol, and solvent extracted materials were tested for inhibition of mutagenicities induced by aflatoxin B 1 (AFB 1), benzo[ a]pyrene (BaP), 2-amino-3-methylimidazo[4,5- f]quinoline (IQ), and cyclophosphamide (CP) in histidine-deficient strains of Salmonella typhimurium. Antimutagenic activities were found in many extracts, but especially in the n-hexane extracts. For example, in the case of oranges, 100 μg of this extract reduced the bacterial mutagenicity of AFB 1, BaP, CP and IQ by 72, 67, 53, and 27%, respectively. Separation by semi-preparative HPLC of the n-hexane extracts of carrots, tomatoes, and oranges indicated that the antimutagenicity was mainly associated with the fractions of the hydrocarbon carotenoids ( α-, β-carotene, lycopene), the xanthophylls ( β-cryptoxanthin, lutein), and also the carotenolesters (oranges). When 16 reference carotenoids were investigated as described above, the following results were obtained: In the case of BaP, antimutagenic activity, quantified by dose–response curves, was exhibited by 8′-apo- β-carotenal, α- and β-carotene, β-cryptoxanthin, lutein, retinal, and retinol (ID 50-values: 20–100 nmol ml −1 top agar, 50–70% maximum inhibition at 1 μmol ml −1 top agar), while the maximum inhibition by torularhodin did not exceed 40%. Astaxanthin, 10′- and 12′-apo- β-carotenal, bixin, canthaxanthin, ethyl-8′-apo- β-caroten-8′-oate, lycopene, and zeaxanthin were inactive or at best marginally active (<20% inhibition). Closely similar results were obtained with AFB 1. The bacterial mutagenicity of CP was strongly reduced by α- and β-carotene, canthaxanthin, and retinol (ID 50-values: 67–112 nmol ml −1 top agar, 50–63% maximum inhibition at 1 μmol ml −1 top agar), moderately by β-cryptoxanthin, and lutein (45% and 28%, respectively), and only marginally or, not at all, by all remaining carotenoids. In the case of IQ, the carotenoids exhibited the weakest antimutagenic potency (7–43%, ID 50-values of retinal and retinol: 160 and 189 nmol ml −1 top agar, 60% and 55% inhibition, respectively). The mutagenic activity of the proximal mutagen of IQ, N-OH-IQ, in S. typhimurium TA 98NR was not significantly reduced by any carotenoid tested. These observations as well as the inhibition of various cytochrome P-450 linked 7-alkoxyresorufin- O-dealkylase activities (EROD, MROD, PROD) by four selected carotenoids (retinol> β-cryptoxanthin> β-carotene>lutein, IC 50-values: 19–109 μM), indicate that the inhibition of the metabolic activation of the different promutagens could cause antimutagenicity. Finally, it could be demonstrated that the number of BaP or CP induced micronuclei in polychromatic erythrocytes in bone-marrow of mice was reduced significantly by the carotenoids lycopene, canthaxanthin, lutein, and β-cryptoxanthin (25–46%). These results clearly show that carotenoids possess biological activities in vitro and in vivo distinct from their function as precursors of vitamin A or antioxidants suggesting effects on activation of promutagens.