RESEARCHhealth promotion ingly in medical and dental research. 6,7 It is particularly suitable for identifying, exploring and explaining complex attitudes, perceptions and beliefs, 8 explaining the level of consensus around a given topic, 9 and can overcome some of the disadvantages of quantitative methods, especially non-sampling error such as the superficiality of response. As Kitzinger put so succinctly:' Focus group discussions involve bringing together, in an informal setting, groups of eight to ten subjects who are carefully selected in social demographic terms, and asking them to discuss areas of interest under the direction of a group moderator. The moderator encourages interaction and synergy between group members and ensures that their subjective experiences are explored in relation to the determined research objectives.Subjects were recruited to take part in one of the focus groups (Table 1) on a door-to-door basis by trained and experienced market research interviewers (a method widely used in social science research and more frequently in health services research) according to a strict code of conduct. 11 Group two comprised members from socio-economic grouping ABC1, group three CDE, and group one was mixed of ABC1/C2DE. Ten subjects were invited for each group; a group comprised a minimum of eight. Both the personal invitation and the incentive ensured a high turn-out. In order to avoid any possible bias, respondents were not informed of the exact nature of the research, but were told that the discussion would be centred around the topic of 'health and illness' . Oral cancer was only mentioned at the end of the discussion if it became apparent that this topic would not arise naturally. Neither dental attendance or personal knowledge of oral cancer were to be recorded. Subjects were recruited on the following basis:• Age (45-60-year-old range) • Male sex • Health status (no mouth or throat problems) • Drinking and smoking habits (either drinkers and/or smokers) • Place of residence (north east of England).The three focus group discussions (Table 1 labelled in order of completion) lasting one and a half hours were held at a central location and subjects were paid a standard fee for participating. The moderator (who conducted and analysed data from all groups) used a standardised guide for the discussion, which included Objective To examine the perceptions and understanding of oral cancer among older male drinkers and smokers in the north east of England. Design Qualitative research using focus group discussions led by an experienced moderator. Setting Residents of the north east of England in their community. Subjects Male alcohol drinkers and tobacco smokers over the age of 44 years and by socio-economic grouping. Results There is a lack of knowledge and understanding of the risk of oral cancer in this whole at-risk population sample. Even those who have direct contact with the disease profess ignorance. Information on health is perceived as confusing or distrusted. Much of this is linked to a fatal...