In this paper, we report on the temperature dependence of electromigration threshold product in copper interconnects. The electromigration threshold product (jL) c is first determined from lifetime distributions for temperatures ranging from 260°C to 330°C. We then propose an alternative method to determine (jL) c at much lower temperatures. We show that the threshold product does not vary significantly in the investigated temperature range. We demonstrate this alternative method is suitable to determine (jL) c close to the final product operating temperature, which would require unrealistic test duration if performed with a conventional method based on lifetime data.