An electric starter provides initial power to run the Auxiliary Power Unit (APU) during the starting process. As the starter degrades, its output power declines which, as a result, affects the APU starting performance, and eventually leads to the starting failure. Therefore, a reliable starter health condition monitoring system is necessary to estimate the starter's degradation, diagnose its fault, and prevent its malfunction. Some research works have been reported in the literature for starter performance estimation and fault diagnostics, however, these prior works typically employ statistical analysis and data-driven methods, which are not physically interpretable and in some cases, are not reliable. Accordingly, the objective of this thesis is to develop a physics-based diagnostic scheme for the starter performance degradation estimation to improve fault diagnostics reliability.