One of the main issues of the high-density, high-radiation plasma-edge scenario for long-pulse operation in future fusion devices is to reduce physical sputtering by lowering the edge electron temperature. If carbon-based materials are used as plasma-facing components in next generation machines such as ignition device to test engineering concepts, the role of chemical with respect to physical sputtering increases as the net erosion mechanism limiting the component’s lifetime. This is due to the fact that the physical sputtering yield decreases stronger when decreasing the ion energy in the 30-150 eV energy range, although very recent laboratory experiments have concluded an important chemical sputtering yield decrease at low deuterium impact energies (down to 10 eV). Because laboratory experiments cannot reproduce the real plasma-wall interaction conditions, principally the high bombarding flux conditions, it is important to determine experimentally and in situ the chemical sputtering yields. The global chemical sputtering yield for a tokamak with carbon walls has been estimated in Tore Supra by analyzing the aftershot molecular outgassing with a differentially pumped mass spectrometer. Quantitative measurements of molecular outfluxes during plasma operation are difficult due to the strong cracking probability of the neutral species. The aftershot outgassing method takes advantage of the fact that just after the termination of the plasma, the molecular desorption responsible for chemical sputtering continues, although in a decreasing way. This avoids the plasma perturbation effect due to molecule cracking. Even though the analysis of the spectrum is difficult due to the complex decomposition pattern of organic compounds containing hydrogen and deuterium, the characteristic time evolution of the different outgassing species can help to deconvolute the mass spectrum. The main product is CD4 and the measured chemical sputtering yield (Y-CD4 = methane outflux from the wall/deuterium flux to the wall) is 3 x 10(-3) < Y-CD4 < 9 x 10(-3) for D-2 discharges. The contribution of second order hydrocarbons and of CO to the total chemical erosion yield is also studied. The dependence of the yields on plasma flux and surface conditions is discussed, as well as the pros and cons of the method used.