1. Isolated kidney cortex tubules prepared from fed rats and incubated with near-physiological concentrations of [(14)C]lactate decrease the specific radioactivity of the added lactate. This effect may be attributable to at least two mechanisms; formation of lactate from endogenous precursors, or entry of unlabelled carbon into the lactate pool as a result of substrate cycling, via phosphoenolpyruvate, pyruvate and oxaloacetate, together with equilibration of the oxaloacetate pool with malate and fumarate. Such substrate cycling could occur within a single cell, or between two populations of different cells, one glycolytic and the other gluconeogenic. These possibilities have been investigated by using metabolic inhibitors or alternative metabolic substrates. 2. Tubules from fed rats produced a fall in specific radioactivity of 14.4% when incubated for 40min with 2mm-lactate alone. A mathematical treatment of this result is presented, which allows the rate of fall in specific radioactivity to be expressed as the addition of unlabelled lactate to the pool. This corresponds to a rate of formation of unlabelled lactate of 121+/-22mumol/h per g dry wt., a rate close to that of gluconeogenesis. In tubules from fasting rats, there was no reduction of the specific radioactivity of lactate, indicating that fasting for 24h suppresses production of unlabelled-lactate carbon. 3. Addition of 2mm-fumarate resulted in a significantly greater decrease in the specific radioactivity of lactate, but aspartate (2mm), malate (2mm) and glucose (5mm) were without effect. Total inhibition of gluconeogenesis with 3-mercaptopicolinate did not prevent the fall in specific radioactivity of lactate observed in tubules from fed-rat kidney, thereby excluding significant activity of the substrate cycle pyruvate-->oxaloacetate-->phosphoenolpyruvate-->pyruvate. 4. The capacity of pyruvate kinase under the test conditions in tubules prepared from kidneys of fed or starved rats was at least ten times higher than the observed rate of production of lactate, so that failure to observe recycling of lactate in starved-rat tubules indicates suppression of pyruvate kinase activity. 5. The endogenous glycogen and glucose content of isolated renal cortex tubules is too low to account for the dilution of label of lactate. Endogenous concentrations of glycerol and amino acids were also very low. As for glycogen, the possibility that very rapid turnover of these metabolites, in fed rats but not in starved rats, may account for formation of unlabelled lactate cannot be excluded. 6. It is concluded that substrate cycling via phosphoenolpyruvate does not occur to any significant extent in either fed or starved-rat kidney. In fed rats recycling of lactate carbon does occur and the rate of this reaction is similar to the rate of gluconeogenesis at physiological concentrations of lactate. The present results favour participation of oxaloacetate decarboxylase rather than ;malic' enzyme in this cycle.