This thesis studies the energy behavior of the Virtual Cathode Oscillators (Vircators). The overall objective focuses on determining geometric and functional parameters to maximize the energy radiated into a specific band of frequency. Initially, the problem was addressed through numerical optimization. A computational tool integrating an evolutionary algorithm with a simulator of particles was developed. The main advantage of this approach is the fact that Vircator of different typologies can be optimized. A second approach focuses on solving the problem through classic optimization techniques. The first step was to determine a mathematical model that relates the Vircator design parameters with the energy output. Then, the mathematical model was studied and optimized. Principal advantages of this approach are the low computational complexity and the fact that the model allows studying and understanding Vircators physics. The approaches presented in this thesis were validated by computational simulation and reports of experiments available in the literature. The main result of this thesis was the identification of the role of the design parameters on the energy response of the Vircators. Additionally, it was found two methodologies to optimize the Vircators’s energy responses at a determined frequency.