Lay beliefs about illness, its causes and its treatment, do not necessarily concur with medical knowledge, and can sometimes be highly idiosyncratic. These beliefs are likely to be influential in help-seeking, in patients' attitudes to professional help, and in the manner in which patients participate is the management of their illness. Clinicians thus need to understand such lay beliefs and attitudes in order to engage their patients in treatment and to provide optimal care. Lay beliefs are likely to be influenced by the individual's culture and hence also by ethnic group. In attempting to understand the patient's beliefs, the researcher or clinician runs the risk of ethnocentricity-viewing the patient's culture inappropriately from the clinician's own perspective. In some senses, this applies to every clinical encounter-patient and clinician always come from different cultures, in the broad sense. Sensitive clinicians develop expertise at bridging this cultural gap and seeing the patient's problems from the latter's viewpoint. However, more systematic investigation of beliefs and attitudes within a given culture can be pursued using the anthropological technique of ethnography. Ethnographic interviewing can yield qualitative data which can then be taken further in quantitative studies. To minimise the risks of ethnocentricity, it may be appropriate to analyse such data not using customary statistical methods but non-linear multivariate data analysis.