The photovoltaic module failure mechanisms related to temperature, humidity, and electrical bias are analyzed using the data collected over a period of 20 years from various sites in the U.S. The approach is based on measuring the rate dependence of the mechanisms on site stress levels, and then using the rate data to analytically estimate the field life by means of computer models of the site environment. A correlation is established between the accelerated constant-stress testing and the time-varying field exposures. Test results are presented for two failure mechanisms for a module design featuring polyvinyl butyral encapsulant for the temperature range of 85 to 100 C and 85-percent relative humidity.