1. The catabolism of purine nucleotides was investigated by both chemical and radiochemical methods in isolated rat hepatocytes, previously incubated with [(14)C]adenine. The production of allantoin reached 32+/-5nmol/min per g of cells (mean+/-s.e.m.) and as much as 30% of the radioactivity incorporated in the adenine nucleotides was lost after 1h. This rate of degradation is severalfold in excess over values previously reported to occur in the liver in vivo. An explanation for this enhancement of catabolism may be the decrease in the concentration of GTP. 2. In a high-speed supernatant of rat liver, adenosine deaminase was maximally inhibited by 0.1mum-coformycin. The activity of AMP deaminase, measured in the presence of its stimulator ATP in the same preparation, as well as the activity of the partially purified enzyme, measured after addition of its physiological inhibitors GTP and Pi, required 50mum-coformycin for maximal inhibition. 3. The production of allantoin by isolated hepatocytes was not influenced by the addition of 0.1mum-coformycin, but was decreased by concentrations of coformycin that were inhibitory for AMP deaminase. With 50mum-coformycin the production of allantoin was decreased by 85% and the formation of radioactive allantoin from [(14)C]adenine nucleotides was completely suppressed. 4. In the presence of 0.1mum-coformycin or in its absence, the addition of fructose (1mg/ml) to the incubation medium caused a rapid degradation of ATP, without equivalent increase in ADP and AMP, followed by transient increases in IMP and in the rate of production of allantoin; adenosine was not detectable. In the presence of 50mum-coformycin, the fructose-induced breakdown of ATP was not modified, but the depletion of the adenine nucleotide pool proceeded much more slowly and the rate of production of allantoin increased only slightly. No rise in IMP concentration could be detected, but AMP increased manyfold and reached values at which a participation of soluble 5'-nucleotidase in the catabolism of adenine nucleotides is most likely. 5. These results are in agreement with the hypothesis that the formation of allantoin is controlled by AMP deaminase. They constitute further evidence that 5'-nucleotidase is inactive on AMP, unless the concentration of this nucleotide rises to unphysiological values.