A current challenge in conducting telephone surveys is obtaining satisfactory representation among younger adults. Surveys using random digit dial (RDD) or directory-listed sample frames routinely encounter coverage issues related to the exclusion of about 25 percent of adults age 18 to 24 who live in households without landline phone service. In addition to this coverage issue, a related problem is the likelihood of contacting and interviewing the 75 percent of young adults who do live in households with landline phone service. Because reliance on cell phones for communication is generally higher among younger adults, young adults who do have landline phones may be less likely to respond to landline phone calls. Telephone surveys that include only landline phones may further under-represent young adults by failing to include those who are difficult to reach by landline phone. To better understand the potential impact of phone usage patterns on nonresponse among young adults, this paper examines patterns of landline and cell phone use among adults age 18 to 24 and assesses the impact of these usage patterns on nonresponse bias. The data are drawn from an RDD-based survey on health behaviors that targeted young adults in the state of New York. All participants were asked about their current phone service and usage patterns with respect to both landline and cell phones. We hypothesized that the young adults interviewed via landline phones would report significantly greater use of cell phones versus landline phones overall. In addition, we expected specific usage patterns could be related to factors such as the number cell phones in the household, sharing of cell phones in the household, and type of housing. Bringing several health indicators into the analyses, we also assessed how young adults’ phone usage patterns could potentially contribute to nonresponse error in RDD surveys on health behaviors.