When wireless telephones are substituted for residential landline telephones, the representativeness of random-digit-dialed (RDD) household telephone surveys is affected. Indeed, demographic differences between wireless-only persons and persons with landline phones have been found by the few studies that have looked for such differences. For example, a study by Mediamark Research Inc. suggested that wireless-only individuals tended to be younger single males, living in rental units in urban areas. Students, unemployed persons, and African-Americans were also more likely to have cut the cord (Giesbrecht & Cynamon, 2002; Piekarski, 2003). Data from the Yankee Group (2003) confirm that younger adults are more likely to have only wireless telephone service, estimating that 12% of 18- to 24-year-old adults have chosen to cut the cord. Given that age, income, and other demographics predict health status and health access, and given that wireless-only persons do differ from landline persons on these demographic characteristics, RDD health surveys that exclude wireless-only persons and that fail to account for this noncoverage bias (e.g., in the creation of the sampling weights) will produce biased estimates. To better understand the implications of wireless substitution for RDD household health surveys, this presentation will highlight data from the 2003 National Health Interview Survey on the health and health care access of wireless-only persons and of persons with landline phones. Then, to determine whether identified biases can be explained by demographic differences, regression analyses will be presented that examine whether a person’s wireless-only status still accounts for variance in health and health care access after controlling for the other demographic characteristics. This analysis will help determine whether standard poststratification techniques to adjust for demographic sample biases will be sufficient to alleviate the noncoverage bias in RDD health surveys that don’t include wireless-only persons.