Livers of rats fasted for 48 h were perfused with [2-14C]fructose, [6-14C]fructose or [U-l4C]-fructose in a non-recirculating perfusion system.From the 14C-labeling patterns of the metabolites glucose and lactate, which were isolated from the effluent perfusate, as well as from the labeling patterns of lactate, phosphoglycerate and phosphoenolpyruvate, which were separated from the tissue extract, the following conclusions were drawn:1. The 14C activity of lactate in the perfusate leaving the liver is almost independent of the position of 14C label in the fructose, which was infused. This indicates that this lactate originates from a metabolic pathway in which both triose units of fructose are converted to lactate. On the other hand the 14C activity of cellular lactate depends on the position of the 14C label in the infused fructose. This lactate seems to be formed only from the C4-C5-C6 unit of the fructose. These two pathways of fructose metabolism are postulated t o be located in different cell compartments of liver tissue and are initiated by phosphorylation of fructose a t carbon position 6 and 1, respectively. We refer to them as the "glycolytic" and "gluconeogenic" compartments.2. I n experiments with [2-14C]fructose there is no correlation between the specific 14C activities of phosphoglycerate, phosphoenolpyruvate and intracellular lactate indicating compartmentation of these metabolites.3. By comparing the specific 14C activities of the C4-C5-C6 triose unit of newly formed glucose (taken as representative of the specific 14C activity of glyceraldehyde-3-phosphate in the "gluconeogenic" compartment) with the specific 14C activity of cellular lactate, the following can be postulated. Glyceraldehyde is transformed into lactate either by a pathway not involving glyceraldehyde-3-phosphate as an intermediate or a glyceraldehyde-3-phosphate which is spatially separated from the intermediate glyceraldehyde-3-phosphate pool of gluconeogenesis. The measured high tissue levels of 2-phosphoglycerate in our experiments support the first concept. 4. I n experiments with [2-14C]fructose, the specific 14C activity of newly formed glucose corresponds to the 14C activity of fructose infused, while in experiments with [6-14C]fructose a high dilution of the 1% activity of glucose was observed. This dilution must originate from a crossing over of some intermediate metabolite in the conversion of fructose to glucose with the inactive intermediate metabolite of some other metabolic pathway. An interaction with the citric acid cycle as possible source of this dilution was excluded.The conversion of fructose to glucose in liver tissue proceeds mainly along a reaction sequence, the first step of which is the phosphorylation of fructose in carbon position 1 [1,2]. There is still some doubt concerning intermediate steps in this Enzymes.