Abstract A model is developed that predicts the shift in distributions of indoor radon concentrations and potential risk reduction resulting from a program of homeowner sampling and remediation in a region. Indoor radon concentrations for a region are represented by a lognormal distribution. Discrete choice functions are used to describe the likelihood of homeowner sampling and remediation. The parameters of these functions are related to government policies for public information and radon risk management. A closed-form equation is derived for the shifted probability distribution function of radon, and this distribution is integrated over the lung cancer dose-response function to estimate the modified population risk in the region. Application of the model is illustrated using radon data from the Reading Prong region in Eastern Pennsylvania and the state of Washington.