The quantity of lactose not absorbed by 4 normal and 6 lactase-deficient subjects was determined by three indirect methods which involved: (1) measurement of pulmonary hydrogen (H2) excretion, (2) pulmonary 14CO2 excretion, and (3) stool 14C excretion, after ingestion of 12.5 g of 1-14O-lactose and 4 g of polyethylene glycol (PEG). Results were compared with absorption determined directly from the 14C:PEG ratio of multiple terminal ileal aspirates., The fraction of lactose not absorbed determined by ileal aspiration ranged from 0 to 8% in normals and 42 to 75% in milk-intolerant subjects. Whereas all three indirect methods were useful in qualitatively separating normal from deficient subjects, the quantity of lactose absorbed as determined by H2 excretion correlated most closely with ileal measurements (r = 0.94). Pulmonary 14CO2 excretion for 24 hr after 14C-lactose ingestion did not distinguish normal (17 ± 4% (SEM) of ingested 14C per 24 hr) from lactase-deficient subjects (21.1 ± 3%). Likewise, stool 14C:PEG ratios grossly underestimated malabsorption with less than one-quarter of the nonabsorbed 14C appearing in the stool. This study suggests that individual differences in susceptibility to diarrhea after milk ingestion by lactase-deficient subjects may be due to differences in the quantity of lactose not absorbed and/or differences in the rate of bacterial metabolism of lactose in the colon. Analysis of ileal fluid collected during passage of the lactose meal indicated that about two-thirds of the osmotic load entering the colon consists of endogenous electrolytes. Thus the water load delivered to the colon is about 3 times that calculated to be osmotically held by the nonabsorbed sugar.