The telephone has enabled significant advances in social science and public health research. Telephone surveys are less expensive, easier to manage, and faster than face-to-face surveys. Nevertheless, survey designers now face different types of populations: landline subscribers—among whom we must distinguish those listed in the phone book—and exclusive mobile phone users (EMPUs). While the telephone remains an excellent means of reaching individuals residing in France (approximately 1% of people have no telephone), the significant proportion of users with unlisted numbers and, above all, of PETPs complicates the methods developed since the early 1990s. Between 1995 and 2005, the proportion of landline subscribers declined in favor of exclusive mobile phone owners. Since then, thanks to the growth of bundled internet-phone packages, landline telephony has experienced a resurgence that has reduced the proportion of exclusive mobile phone owners. In the context of the last two editions of the INPES Health Barometer, the PETP sample was selected independently of the landline sample. The telephone number prefixes assigned by the French Regulatory Authority for Electronic Communications and Postal Services (ARCEP) to the various national operators were used to stratify the sample. The remaining digits of the numbers were then generated randomly. Furthermore, following the opening of the telecommunications market in 2003, the incumbent operator was required to grant access to the fixed-line network to other operators. The possibility of full unbundling thus gradually led to a loss of subscribers for France Télécom. According to ARCEP, the number of lines under full unbundling, which was less than 100,000 (or 0.4% of households) in early 2005, has continued to grow since then, reaching more than 5 million households (or approximately 15% of households) in 2009. The actual ability to keep one’s old geographic number, even when switching carriers, became available fairly quickly after the opening of the telecommunications market, so that households reachable only via a number starting with 08 or 09 represented a small share of the total population (2.5%) in 2008, though this figure appears to be rising. For a single telephone line, one may therefore have a geographic number and/or a number starting with 08 or 09. Due to the impossibility of managing duplicate counts between geographic numbers and numbers starting with 08 or 09, it was not possible to include the latter in the landline survey frame. Nevertheless, it seemed unjustified to exclude them outright from the surveyed population, given the lack of information on this group. This precaution required an adjustment to the survey method to identify those “unbundled to 08 or 09” via the PETP. In fact, nearly all (approximately 95%) of the total unbundled users who report that they cannot be reached at a geographic number own a cell phone. Thus, thanks to a screening question asked of the PETP sample, it was possible to identify households reporting that they also have a landline but do not correspond to any geographic number. Ultimately, the survey’s scope leaves out almost only those individuals who do not have any telephone equipment. [author’s summary]