Rice starch is a main source of energy in many lesser developed countries. We studied different chain-lengths of rice glucose polymers (GP) to evaluate their possible use in feeding infants in developing countries. The initial GP of rice (G1 = 4.6, G2 = 4.5, G3 = 15.4, G4 = 7.3, G5 = 17.4, G6-G9 = 9.61 and greater than G9 = 31.3%) was analyzed by HPLC and then separated in a Bio-Gel P-2 column and compared to its short-chain GP of rice (G2 = 22.7, G3 = 28.2, G4 = 14.0, G5 = 16.6, G6 = 11.6, G7-G9 = 6.9%), long-chain GP of rice (greater than G9 = 100%), and D-glucose. Intraduodenal bolus infusion of 10% solution of short-chain rice GP when compared with long-chain rice GP, the initial rice GP, or D-glucose showed significantly higher values at peak absorption time (0 to 30 min) in the portal venous blood glucose response. The portal venous glycemic response of short-chain rice GP compared with D-glucose was as follows: 2.5 +/- 0.1 versus 2.0 +/- 0.2 cm2, area under the portal blood glucose curve at 0-30 min (p less than 0.01). Glucoamylase, the key enzyme for brush-border hydrolysis of short-chain GP, was assessed with a newly modified glucoamylase assay using GP G5-G8 as substrate. Our finding of faster glucose absorption with short-chain rice GP compared with isocaloric D-glucose might have important physiologic implications for carbohydrate absorption. The osmolality of short-chain rice GP is nearly one-fourth that of glucose. This might have important bearing in the design of infant feeding where increased caloric density with low osmolality is desirable.