Master of Science / Department of Electrical and Computer Engineering / Behrooz Mirafzal / Increasing energy demand, rising oil prices, and environmental concerns have forced attention to alternative energy sources that are environmentally friendly and independent of fossil fuels. Renewable energy sources (RES) have become an attractive alternative to the traditional energy sources for electric power generation. However, one of the main challenges of RES adaption arises when connecting RES to the electric grid. Voltage source inverters (VSIs), typically, connect RES to the electric grid. Similar to any engineering system, detailed dynamic models of the VSIs are needed for design and analysis purposes. However, due to the non-linearity of VSIs, development of dynamic models that can accurately describe their behavior is a complex task. In this thesis, a detailed averaged-state-space model of the two-level three-phase space vector pulse width modulation VSI and its companion LCL filter is derived. Because VSIs can operate under stand-alone and grid-tied modes, two models were derived for each case. In the derived models, the VSI modulation index m and phase angle ϕ are initially considered constant. In practice, however, these parameters are considered the main control parameters. To model these parameters as control inputs, small-signal models of the VSI under stand-alone and grid-tied modes were derived. To verify the accuracy of the developed large-signal and small-signal models, Matlab/Simulink simulations were carried out. The simulation results were compared against the models results. Moreover, the models were verified through lab experiments. The developed models can be used as design and analysis tools. In addition, the developed models can be used as fast and efficient simulation tools for system studies, when the modeling of switching transients is not needed. Nowadays, the number of VSIs connected to the electric grid is growing exponentially. The amount of time and computation needed to simulate VSIs using simulation software packages can be significantly decreased by the use of the developed models.