There is a paucity of research on health in the UK Chinese community partly due to the difficulties of identifying and accessing study populations. For a survey of cardiovascular disease we aimed to identify and recruit all Chinese adults aged 25-64 y living in Newcastle-upon-Tyne, UK. One thousand, eight hundred and sixty-five potential subjects were identified using a variety of methods. Of the 1702 potential subjects identified from a name analysis of the 1991 FHSA register (FHSA group), 638 students in halls of residence were excluded and the remaining 1064 were invited to participate. Non-respondents were followed up. Of the 1064, 658 (65.5%) addresses were no longer valid, 21 (2%) were reclassified as non-Chinese and no contact was made with 18 individuals (1.6%). A further 163 subjects (non-FHSA group) came forward in response to publicity, giving a total of 530 Chinese actually identified in Newcastle. Three hundred and eighty subjects took part in the study. Compared to the 1991 Census, the recruitment procedure underestimated the total population size, particularly for men and younger ages. In the FHSA group, men were significantly more likely to be current drinkers, and women were more likely to smoke and have a lower educational attainment that the non-FHSA group. There were no other important differences in the distribution of CHD risk markers in the two groups. Our experience indicates that the FHSA register is suitable for identifying Chinese but should be used alongside other complementary methods to augment samples for ethnicity and health research.