This paper presents the application of signomial programming to the assessment and multidisciplinary optimization of electric propulsion systems. Analytic models for electrical and mechanical propulsion system components are developed that are compatible with the signo-mial programming objective and constraint forms. These models capture the dependence of mass and efficiency on material properties and operating conditions. Standalone optimization of the electrical cable and motor model illustrates a trade-off between mass and efficiency at the component level. These component models are subsequently integrated into various propulsion architectures and are optimized while taking into account the component dependencies (e.g., fan and motor shaft speed). Specifically, turbofan, turboelectric and geared turboelectric propulsion architectures are optimized on the metrics of fuel consumption and propulsion system mass. The results show that the optimal component masses and efficiencies depend on the propulsion architecture, operating point, and performance metric under consideration, and that signomial programming is useful for determining them.