Objectives: To assess the public's knowledge and attitudes to antibiotics, their reported antibiotic use and the relationship between them. Patients and methods:A questionnaire was included in the face-to-face Office for National Statistics Omnibus Household Survey in Britain in 2003. Of 10 981 randomly selected adults from England, Scotland and Wales, 7120 (65%) completed the questionnaire. Results: Although 79% of respondents were aware that ‘antibiotic resistance is a problem in British hospitals’, 38% of respondents did not know that antibiotics do not work against most coughs or colds and 43% did not know that ‘antibiotics can kill the bacteria that normally live on the skin and in the gut’. Respondents with lower educational qualifications were less knowledgeable about antibiotics. In a multivariable analysis, better knowledge of antibiotics was not associated with being less likely to be prescribed any in the last year, but was independently associated with being more likely to finish a course of antibiotic as prescribed. Knowledge was also associated with being more likely to take antibiotics without being told to do so. In women, better knowledge was associated with being more likely to give an antibiotic to someone else that was not prescribed for them. Conclusions: We have shown that there is no simple relationship between increased knowledge and more prudent antibiotic use. Future national antibiotic campaigns should have a defined audience and aims in order to facilitate prudent antibiotic use by clinicians and public.