This proof of concept study is dedicated to the quantification of the short-term recovery phase of the muscle oxygenation and whole-body oxygen uptake kinetics following an exhaustive cycling protocol. Data of 15 healthy young participants (age 26.1 ± 2.8 years, peak oxygen uptake 54.1 ± 5.1 mL∗min-1∗kg-1) were recorded during 5 min cool down-cycling with a power output of 50 W on an electro-magnetically braked cycle ergometer. The oxygen uptake (VO2) signal during recovery was modeled by exponential function. Using the model parameters, the time (T1/2) needed to return VO2 to 50% of VO2peak was determined. The Hill’s model was used to analyze the kinetics of oxyhemoglobin concentration (Sm, %), non-invasively recorded by near-infrared spectroscopy (NIRS) over the M. vastus lateralis. Analysis of the Pearson correlation results in statistically significant negative relationships between T1/2 and relative VO2peak (r = −0.7). Relevant significant correlations were determined between constant defining the slope of VO2 decrease (parameter B) and the duration of the anaerobic phase (r = −0.59), as well as between Hill’s coefficient and average median Smmax for the final 2 min of recovery. The high correlation between traditional variables commonly used to represent the cardio-metabolic capacity and the parameters of fits from exponential and Hill models attests the validity of our approach. Thus, proposed descriptors, derived from non-invasive NIRS monitoring during recovery, seem to reflect aerobic capacity. However, the practical usefulness of such modeling for clinical or other vulnerable populations has to be explored in studies using alternative testing protocols.