As the size of the U.S. population that lives in households with no landline telephones has burgeoned, Arbitron has undertaken a plan of study aimed at finding methods to mitigate potential bias due to frame noncoverage of this population in their radio audience measurement surveys. This plan of study encompasses a number of different facets, including an examination of the cost-effectiveness of sampling the cell-phone-only (CPO) population. This paper reports the results of statistical analyses designed to study the characteristics of the CPO population and how the information gleaned from these analyses might be used to improve the quality of Arbitron’s audience measurement estimates in the face of the CPO issue.