Ground surface analysis of CO2 emissions with δ13C determination is experimentally demonstrated to be a potential methodology to monitor, on line, the dynamics of petroleum-hydrocarbon biodegradation in soil aquifers, thanks to the improvement of the Isotopic Ratio Infra Red Spectroscopy technique. Biodegradation rate of remaining hydrocarbon substrates in groundwater can be quantified using basic application of the Rayleigh equations, by δ13CCO2 analysis released at ground surface above the pollution plume instead of usual approaches based on groundwater hydrocarbons δ13C analysis, when physical and chemical properties for the contaminated site meet appropriate conditions. The validation approach for that gasoline contaminated specific site is discussed and verified by comparison of first order attenuation rate constant determined from δ13CCO2 analysis emitted at ground surface and from δ13CTOLUENE analysis in ground water. A kinetic fractionation factor α of 0.9979 (or ε value of −2.1 ± 0.5‰) is estimated for the biodegradation of the most reactive hydrocarbon substrates (TEX). The treatment of this Rayleigh equations by linear regression of δ13CCO2 values along the predominant direction of groundwater flow leads to the following results and conclusions for that site: (i) first order biodegradation rate constants (and annual variation) are maximum after the activation of a Permeable Reactive Barrier (PRB) in May 2014: 0.92(+0.29–0.17) year−1, and during July and October: 0.46(+0.14–0.09) year−1 and minimum in mid-winter in February 2015: 0.17(+0.05–0.03) year−1, given by the estimation range for ε. These results are in the lower range with reported in literature for similar contaminated sites (1.6–18 year−1) considering natural attenuation under sulfate reducing conditions and (ii) the seasonal variation of the first order biodegradation rate constant is mainly correlated with the seasonal variation of the CO2 flux, where maximum values are in summers and minimum values in winters. Both seasonal variations are mainly due to the annual cycle of the natural biodegradation activity at the scale of the pollution plume, rather than the activation of the PRB. This work demonstrates that δ13CCO2 analysis released at ground surface from biodegradation of groundwater hydrocarbons could provide, under characterized and appropriate conditions, a non-intrusive (without soil samplings), fast, and low-cost online method to monitor and therefore to optimize soil remediation processes in real time. (Monitored Natural Attenuation or Enhanced Bioremediation).