In this dissertation, we analyzed selected deposition and reaction processes on Si(100) using Kinetic Monte Carlo (KMC) simulation of suitable atomistic lattice-gas models. For the first project on the room-temperature deposition of Ga on Si(100), analysis of the island size distribution reveals that it is monotonically decreasing rather than monomodal, a clear departure from classical nucleation theory. KMC simulations of a realistic model recover this behavior, and simulations of refined models show that aggregation is responsible for the unexpected shape of the island size distribution. From these simulations, we were also able to extract reasonable values for the anisotropic diffusion barriers which recover key aspects of experimental data (e.g. average island size). Next, we extended this analysis to other known systems which exhibit a monotonically decreasing form of the island size distribution and found a large ratio of the nucleation rate to the aggregation rate in all cases.;For the second project, we studied oxidation and etching of oxygen on Si(100) for a wide range of temperatures ranging from high to moderate. In the high temperature regime, etching via the removal of SiO dominates but increasingly competes with the formation of oxide clusters as the temperature approaches the 550 C--700 C range. After prolonged etching, oxide-capped Si nanoprotusions gradually emerge as the dominant feature of the surface. We first developed a simplified atomistic model to describe this behavior under the conditions of layer-by-layer etching of a flat surface. In the subsequent modeling, focus was on oxidation and etching of vicinal Si(100) where detailed treatment of the equilibrium step structure and Si(100) dynamics makes it possible to analyze phenomenon such as step recession, pinning of receding steps by oxide clusters leading to the creation of "fingers", their subsequent pinch-off, and break-away islands. To our knowledge, this was the first attempt to include detailed description of the key surface diffusion processes (like di-vacancy diffusion) in lattice-gas modeling. In addition, we have also explored the effects of an enhanced nucleation at the step edges and its role in the emergence of fingers.