Critical issues regarding the microwave pulse induced irreversible failures in individual metal–oxide–semiconductor field‐effect transistors and CMOS integrated circuits have been well addressed in this article. The fundamental mechanisms for irreversible failures are analyzed, and a theoretical foundation based on the microwave pulse induced thermal effects is developed. Two theoretical models that capture the effects of pulse width and pulse repetition frequency (PRF), respectively, are formulated. In comparison to the numerical results obtained from device simulations and the experimental data reported in literature, the validity of these two theoretical models has been verified. The pulse width dependent failure model describes an asymptotic behavior of a failure threshold versus pulse width dependence. It suggests a substantial decrease in the failure threshold for very short pulse width range and a smooth transition to constant power requirement for long pulse width range due to thermal equilibrium. The PRF dependent failure model indicates an inverse relationship between the failure threshold and the PRF or the number of pulses. © 2014 Wiley Periodicals, Inc. Microwave Opt Technol Lett 56:1144–1151, 2014