Plasma samples were obtained before, and 8 and 24 h after the ingestion of an ad libitum amount of cooked carrots by 23 school children from a peri-urban community in Guatemala City. The single-meal consumption of cooked carrots ranged from a low 122 g to a high of 961 g (mean: 370.5 +/- 237.2 g; median: 268.4 g). The measured beta-carotene content of the carrots was 10.1 mg per 100 g of edible portion; therefore, the range of intake of beta-carotene was 12.4 to 97.0 mg (mean: 37.4 +/- 24 mg; median: 27.1 mg). Changes in plasma beta-carotene levels at 8 h ranged from a decrement of -0.32 mumol/l (-16.98 micrograms/dl) to an increment of 0.79 mumol/l (42.44 micrograms/dl), with a mean of 0.11 +/- 0.24 mumol/l (5.97 +/- 12.82 micrograms/dl). Changes at 24 h were less dramatic than those at 8 h. A regression of the 8-h changes in plasma beta-carotene on the amount of carrot carotene consumed (corrected by body weight) had an r-value of 0.12. Baseline levels of plasma retinol were poor predictors of the plasma beta-carotene response with this sample size (r = 0.10). The magnitude of the plasma response to beta-carotene from carrots appears to be lower than that observed with pure, powdered, crystalline carotenes; moreover, the variability of the post-carrot response seems to be greater--and its association to dosage appears to be weaker--than with the pharmacological beta-carotene.