Traditionally, it has been accepted that working muscle will endeavor to maintain [ATP] near its resting level in order to maintain a favorable free energy of ATP hydrolysis (∆GATP). Recent reports of significant declines in [ATP] during maximal muscular contractions in vivo spark new interest in this concept. ATP is hydrolyzed to adenosine diphosphate (ADP) in the creatine kinase reaction to provide energy directly to working ATPases. As needed, additional energy can be provided by the breakdown of ADP to adenosine monophosphate (AMP), which is rapidly cleared by the formation of inosine monophosphate (IMP), with accompanying loss of total adenine nucleotides. IMP formation has been thought to occur only during very intense muscular work, wherein ATP production does not keep pace with ATP consumption. Here, we propose a role for IMP in managing both ∆GATP and flux through a key energetic pathway during contractions. To test our hypothesis that IMP formation mitigates the accumulation of ADP, 8 males (28.4 + 3.5 yr, mean + SD) performed 8 min of incremental, isotonic knee extension contractions (1 every 2 s, starting at 6% peak torque and incrementing by 3% every 2 min) on a magnetic resonance (MR)‐compatible ergometer in a 3‐Tesla MR system. 31Phosphorus MR spectra were obtained at rest (60‐s average) and throughout the contraction protocol (10‐s temporal resolution). Peaks corresponding to ATP, phosphocreatine (PCr), inorganic phosphate (Pi), and phosphomonoesters (PME) were fit in jMRUI, and concentrations for these metabolites were determined assuming resting [ATP] = 8.2 mM. [ADP], [AMP], [IMP], and the rate of ATP (mM∙s‐1) production by non‐oxidative glycolysis (i.e., wherein the fate of pyruvate is lactate; ATPgly) were also calculated. As expected, [PCr] decreased and [Pi] increased approximately linearly throughout the incremental contraction protocol. Surprisingly, [ATP] decreased and [IMP] increased (to 7.3 + 0.4 mM and 0.8 + 0.56 mM, respectively) steadily from the onset of contractions until the final stage, at which point no further changes were observed. In contrast, both [ADP] and [AMP] increased linearly from the start of contractions until the final stage, when they accumulated more rapidly; ultimately to 109.9 + 52.1 uM and 1.0 + 0.91 uM, respectively. The glycolytic contribution to ATP production was consistent for the first 3 stages, after which ATPgly increased rapidly to 0.102 + 0.14 mM·s‐1 at 480 s. Collectively, the timing and magnitude of these metabolic changes in working muscle suggest a sequence of responses designed to maintain cytosolic [ADP] sufficiently low as to mitigate changes in ∆GATP, but at a level sufficient to stimulate ATPgly as contraction intensity increases. A novel observation in this study was the decrease in [ATP] (and thus increase in [IMP]) from the onset of the contraction protocol, which then leveled off at the highest contraction intensity. The utility and consequences of these responses remain to be determined, but this study supports the concept that IMP formation is important in maintaining the cellular energy state as well as in directing the bioenergetic response to a series of incremental muscular contractions in vivo.