This study is the first attempt to measure the prevalence of problem gambling attributable to a specific sector of the gambling industry. One thousand, one hundred and five casino patrons in 40 casinos in the UK were interviewed, face-to-face. Respondents were screened for problem gambling using a multiple response version of DSM-IV (DSM-IV-MR). The study found support for Eadington's (1988) hypothesis, that UK casinos could be largely sustained by regular players, among whom the prevalence of problem gambling is high. The study also found support for the hypothesis that, to the extent different gambling sectors are patronised by demographically different client groups, so the problem gamblers associated with them will reflect these client groups. The problem gamblers among the regular casino patrons were demographically distinct from the problem gamblers in the sample who showed a preference for other gambling forms. Other key findings support those found in other jurisdictions. Sector-specific prevalence studies may be the next step forward in epidemiological research on problem gambling. They have the major advantage of netting significantly more problem gamblers from much smaller samples than similar studies in the general population. They also have the potential to reveal the proportion of problem gamblers attributable to each sector, along with their demographic characteristics. Such information would result in more specific information being available for regulators seeking to minimise the social impact of problem gambling and those involved in the development of prevention and treatment strategies.