This paper presents a comprehensive computational analysis of the spectral optical response of epithelial tissues labeled with gold nanoparticles. Monte Carlo modeling is employed to simulate nanoparticle-induced changes in reflectance signals and to assess whether labeling can generate sufficient exogenous contrast that can better pinpoint precancer progression. Simulation results suggest that the observed contrast profile is highly dependent on a series of factors including the labeling scheme, optical sensor geometry, and wavelength. It is evident, however, that selection of an optimal labeling and sensing strategy can lead to a significant enhancement of the inherent positive or negative contrast and can improve the diagnostic potential of optical measurements.