According to one embodiment of the invention, a method for predicting burn-in conditions includes identifying a baseline IDDQ, a baseline temperature, and a baseline IDDQ current density based on a plurality of existing burn-in data for one or more existing devices, determining a theoretical IDDQ current density for a device, determining a ratio of the theoretical IDDQ current density to the baseline IDDQ current density, determining a theoretical process metric for the device at the baseline temperature based on the ratio and the baseline IDDQ, measuring a process metric for an actual device, comparing the process metric for the actual device and the theoretical process metric for the device, and determining an actual burn-in temperature for the actual device based on the comparison.