The high temperature engineering thermoplastic poly(aryl-ether–ether-ketone) (PEEK) has been subjected to dynamic and isothermal thermogravimetry in both nitrogen and air. The dynamic data have been analyzed using both the Kissinger peak maximum technique and an isoconversional procedure developed by Flynn. These techniques gave apparent global activation energies of 223.5 and 235.7 kJ/mol, respectively, for the degradation of PEEK in nitrogen, in good agreement with the value of 219.7 determined from the isothermal experiments. The thermal stability of PEEK in air is substantially less than in nitrogen, and the decomposition mechanism is more complex. The global apparent activation energies for the weight loss in air were found to be 116.9 and 159.5 kJ/mol from dynamic slow heat rate data and isothermal data, respectively. The data obtained from fast heating rate experiments in air were found to be misleading, suggesting caution in the use of “commercial” software packages for lifetime estimates under these conditions, especially where oxidative processes may be occurring.