1. The heat production and mechanical responses of frogs' sartorii have been recorded at 0 degrees C after immersion in normal Ringer solution, and also after poisoning with 1-fluoro-2, 4-dinitrobenzene (FDNB) and nitrogen. The muscles were later analysed chemically for changes in ATP, phosphocreatine (PC), inorganic phosphate (P(i)), lactate, total adenine nucleotide and total inosine nucleotide.2. Analysis of paired resting muscles established that the resting levels of these substances found in our experiments were similar to those reported elsewhere.3. Resting muscles that had been poisoned with FDNB and N(2) contained significantly more PC and less ATP than unpoisoned controls. Moreover, some of their adenine had been deaminated to inosine.4. In a normal muscle in oxygen, the PC that breaks down as a result of a 30 sec tetanus is restored with a roughly exponential time course whose half-time is about 10 min. Thus at least 40 min rest must be allowed between the different stages of an experiment.5. Isometric twitches of the poisoned muscle rapidly decline in size, but small twitches continue to be produced for a very long time. If stimulation is discontinued, substantial recovery takes place. The total tension development is equivalent to at least thirty normal twitches, and correspondingly, the total heat production is greater than could be accounted for even by complete break-down of the ATP in the muscle. In fact, the ATP break-down, though highly significant, is not nearly complete.6. In short series of isometric twitches there is significant break-down of ATP and, less consistently, of PC; also a significant increase in P(i) but no additional deamination of adenine. The rate of development of isometric tension is slightly decreased.7. The results described under (6) are definitely due to the presence of FDNB. In similar experiments with the muscles in N(2) but not otherwise poisoned the only significant change is a break-down of PC.8. In a long series of isotonic twitches the chemical effects are similar to those described under (5) and inosine formation is clearly demonstrable. The poisoned muscles show force-velocity curves of normal shape, the Fenn effect and the presence of shortening heat. Thus their contractions are normal or nearly so.9. The complexity of the chemical reactions that continue in the poisoned muscle makes it impossible to draw up an accurate energetic balance sheet.10. A new hypothesis is suggested to explain the energetic importance of inosine formation.