We have analyzed by (31)P MRS the relationship between kinetic parameters of phosphocreatine (PCr) recovery and end-of-exercise status under conditions of moderate and large acidosis induced by dynamic exercise. Thirteen healthy subjects performed muscular contractions at 0.47 Hz (low frequency, moderate exercise) and 0.85 Hz (high frequency, heavy exercise). The rate constant of PCr resynthesis (k(PCr)) varied greatly among subjects (variation coefficients: 43 vs. 57% for LF vs. HF exercises) and protocols (k(PCr) values: 1.3+/-0.5 min(-1) vs. 0.9+/-0.5 min(-1) for LF vs. HF exercises, P<0.03). The large intersubject variability can be captured into a linear relationship between k(PCr), the amount of PCr consumed ([PCr(2)]) and pH reached at the end of exercise (pH(end)) (k(PCr)=-3.3+0.7 pH(end)-0.03 [PCr(2)]; P=0.0007; r=0.61). This dual relationship illustrates that mitochondrial activity is affected by end-of-exercise metabolic status and allows reliable comparisons between control, diseased and trained muscles. In contrast to k(PCr), the initial rate of PCr recovery and the maximum oxidative capacity were always constant whatever the metabolic conditions of end-of-exercise and can then be additionally used in the identification of dysfunctions in the oxidative metabolic pathway.