In recent years, 210Po implanted in glass artefacts has been used as an indicator of the mean radon gas concentration in dwellings in the past. Glass artefacts have been selected in many dwellings and the alpha-recoil implanted 210Po concentration has been measured using various techniques. Some of these retrospective techniques use a model to estimate the retrospective radon gas on the basis of this surface 210Po activity. The accumulation of 210Po on glass surfaces is determined by the deposition regime over the exposure period. The 210Po activity is determined not only by the radon progeny deposition velocities, but by other room parameters such as ventilation rate, aerosol conditions and the surface to volume ratio of the room. Up to now in using room models, a nominal or 'base-case' scenario is used, i.e. a single value is chosen for each input parameter. In this paper a Monte-Carlo analysis is presented in which a probability distribution for each parameter is chosen, based on measurements quoted in the literature. A 210Po surface activity is calculated using a single value drawn from each of the parameter distributions using a pseudo-random number generator. This process is repeated n times (up to 20,000), producing n independent scenarios with corresponding 210Po values. This process permits a sensitivity analysis to be carried out to see the effect of changes in inputs on the model output.