Power system stability has always been a serious issue for system operators. Power System Stabilizer (PSS) is a common equipment to improve stability conditions in generators. PSS improve system stability by shifting eigenvalues of system. A common method in traditional PSS is to presume grid parameters, as well as load characteristics as constant values. The fact is, however, in real operating conditions, system and load parameters are not constant. Therefore, different variables and uncertainties such as load value, network design and changes in controller parameters have to be considered. Proper statistic distribution is a common method for modeling uncertainties of parameters. This study focuses on designing PSS via analysis of probabilistic eigenvalues. Then, to prove design authenticity, the method is applied on a test 8-machie system.