In view of controversies about assessment of the β-carotene cleavage activity, methodological aspects and problems of the dioxygenase assay are described. Using rat and hamster intestinal preparations the method was optimized on retinal formation, the only cleavage product we could demonstrate. It appeared that the cell fraction with the highest cleavage activity was the 9,000g supernatant (S-9). Maximal retinal formation was obtained with SDS, taurocholate and egg lecithin in the buffer and 3 μg β-carotene dissolved in acetone. Ethanol, THF/DMSO (1:1) or propylene glycol as solvent for β-carotene reduced retinal formation to 55, 24, and 19%, respectively. Retinal formation increased proportionally with the amount of protein S-9 used and was linear up to 40-60 minutes of incubation. Incubation with α-carotene or β-cryptoxanthin resulted in a retinal formation of 29 and 55% of the amount formed from β-carotene. Addition of 9 μg of lutein to an incubation with 3 μg β-carotene reduced retinal formation, while lycopene had no effect. In conclusion, the β-carotene cleavage assay with S-9 as enzyme source described in this report, seems a useful tool to study (dietary) determinants of β-carotene cleavage activity, but for other purposes adaptation of the method is required. Chemicals/CAS: alpha-carotene, 432-70-2; Bcdo protein, rat, EC 1.14.99.36; beta Carotene, 7235-40-7; beta-Carotene 15,15'-Monooxygenase, EC 1.14.99.36; Carotenoids, 36-88-4; cryptoxanthin, 472-70-8; lycopene, 502-65-8; Oxygenases, EC 1.13.-; Retinaldehyde, 116-31-4; Tritium, 10028-17-8; Xanthophylls