Recent investigations have,shown that resistivity decay in evaporated gold films originates, in part, from the annihilation of excess vacancies, trapped during film condensation, at a high density (1015-1{~17/cm 8) of small (~50A) voids present within the film. This paper presents a theoretical treatment of vacancy diffusion to free surfaces and to voids. The treatment predicts both the observed void growth and resistivity decay in thin films.It is well known that the resistivity of freshly deposited thin films changes continuously with time at room temperature (1-3). It has been generally accepted that resistivity decay in thin films takes place as a result of the diffusion of excess vacancies, quenched-in during film formation, to sinks such as free surfaces and grain boundaries (1, 3). Recent transmission-electron-microscope (TEM) studies of evaporated gold films have shown that thin films generally contain a high density (~1016/cm 3) of small voids (4, 5). The presence of such a high density of voids suggests that these voids may serve as potential sites for the annihilation of excess vacancies. Experimental 'evidence for this suggestion has been found in our recent observations that these voids grow at the expense of excess vacancies with aging time even at room temperature (6, 7). Assuming that voids are the only sinks for excess vacancies, Lloyd and Nakahara (6) were able to relate, semiquantitatively, the void growth to the resistivity decay in evaporated gold films. Although the effect of the surfaces was not considered, this treatment was in good agreement with experiments for long times following film deposition, but could not account for short-term resistivity decay. We have, therefore, treated our earlier problem by using a more rigorous mathematical framework which considers both surfaces and voids as potential sinks for excess vacancies.