The (31)P NMR spectrum of energy-related metabolites under strictly aerobic conditions in rapidly respiring tissues under physiological conditions has been approached by the study of the (31)P NMR signals in vivo and in freeze-trapped organs. Freezing the head of the anesthetized animal by liquid N(2), excision of the brain tissue (white and gray matter) at -196 degrees , and transfer to the NMR tube occurs without alteration of the metabolite concentrations. The sample is warmed to the region -15 degrees to -10 degrees , at which temperatures there is sufficient mobility for recording (31)P NMR at concentrations characteristic of brain tissues ( approximately 5 mM) with an adequate signal to noise ratio in 10 min but insufficient mobility for significant enzymatic activity. A approximately 0.4-sec acquisition time is adequate for nuclear relaxation and a 10-min scan gives an adequate signal to noise ratio. Metabolism of creatine phosphate, P(i), and sugar phosphates occurs by 1 hr at -10 degrees and 2 hr at -12 degrees . Extrapolation of the approximately zero order kinetics of disappearance of creatine phosphate and appearance of P(i) suggests that <10% of these two metabolites has been altered in the time of the first measurement.A comparison of the freeze-trapped state and the in vivo metabolite pattern is afforded in preliminary experiments on the head of the living mouse (brain and skeletal tissue) in aerobic and anaerobic states. Longer relaxation times and mild hypoxia due to the restricted diameter of the NMR tube gives significantly lower creatine phosphate/ATP values for this condition. Both direct in vivo and freeze-trapped assays of energy-related metabolites afford excellent approaches to the detection of anoxia and to the evaluation of metabolic control in hypoxic conditions.