The extent to which processing affects the carotene or vitamin A value of foods is poorly understood. An extrinsic reference method was used to estimate the mass of carotenes and vitamin A derived from various preparations made from the same lot of carrots. Using a repeated-measures design, nine healthy adult subjects consumed test meals of either carrot puree (commercial baby food) or boiled-mashed carrots on separate days; six of the subjects also consumed a test meal of raw-grated carrot. Test meals supplied 34.7 micromol (18.6 mg) carrot beta-carotene (beta C), plus 6 micromol deuterium-labeled retinyl acetate (d(4)-RA) in oil solution. Baseline-adjusted carotene and retinyl ester (R-ester) area-under-curve (AUC) responses in the triacylglycerol-rich lipoprotein (TRL) fraction (0-8.5 h) were determined using HPLC and gas chromatography-mass spectrometry. The masses of absorbed beta C, alpha-carotene (alpha C) and R-ester were estimated by comparing their AUC values with that of deuterium-labeled retinyl ester (d(4)-R-ester), assuming the latter represented 80% of the d(4)-RA reference dose. Absorption of beta C and alpha C was approximately twofold greater from carrot puree than from boiled-mashed carrots, whereas the retinol yield was only marginally (P = 0.11) influenced by treatment. Carotene and R-ester absorption from raw-grated carrot was intermediate to, and did not differ significantly from the cooked preparations. The vitamin A yield (puree, 0.53 mg; boiled-mashed, 0.44 mg) of cooked carrot containing 18.6 mg beta C was substantially less than that predicted by current convention and limited primarily by intestinal carotene uptake. Processing can therefore significantly improve bioavailability of carrot carotenes, and in some cases influence the carotene value more than the intrinsic vitamin A value.