Abstract: In many-body systems, there is a strong need to more accurately describe electronic structure. Quantum Monte Carlo (QMC) is a very promising approach to model properties of many-body quantum systems in both the ground state and finite temperature equilibrium. This thesis presents techniques for simulation of semiconductor nanostructures and molecules within Feynman's path integral formulation of statistical mechanics. At low temperatures and large particle numbers, fermionic simulations show an exponentially-growing inefficiency, called 'the fermion sign problem.' To overcome this difficulty we developed a fixed-phase approximation that uses density matrices. With the fixed-phase approximation, the difficulties created by phases in path integrals for magnetic systems are also managed in a practical way. The method is applied to several systems: parabolically-confined fermions in a magnetic field, a realistic three-dimensional model of an InGaAs/GaAs lens-shaped self-assembled dot, and free fermions in periodic boundary conditions. A continuous-space path integral method was also developed for spin 1/2 fermions using the density matrix based fixed-phase approximation. We also apply Path integral Monte Carlo with correlation function analysis to sample quantum mechanical properties of molecules at finite temperature. To accurately treat Coulomb interaction in the many-body systems, new technique for fast and accurate tabulation of the imaginary-time Coulomb Green's function was developed, which is important for quantum simulation of Coulomb systems.