The assessment of long-term isolation performance for a geologic repository requires the use of mathematical models that consider the probability and consequences of postulated disruptive scenarios. In the case of the proposed repository at Yucca Mountain, Nevada, volcanism is one of the important disruptive scenarios being considered in site evaluation. A stochastic modeling approach is developed for use in simulating the airborne release of radioactive particulates associated with the basaltic volcanism scenario. The modeling approach considers such factors as the eruption energetics, eruption duration, wind velocity, and particle properties to compute the activity areal density as a function of spatial location. Various components of the model are based on empirical relationships and data that are reported for observed and monitored cinder cone eruptions analogous to those that likely occurred in the Yucca Mountain region in the past. Illustrative applications of the stochastic model are presented for the cases of a single-event realization and a multiple-event average realization.