The aim of this study was to determine the level of usage of National Health Service Direct (NHSD) by ethnic and gender groups within an urban population. The study population comprised all individuals in the City of Preston, Lancashire, United Kingdom, who had used NHS Direct between 01 July 2003 and 31 December 2003. The ethnic and gender composition of this population was compared with that of the total population within the study area. The United Kingdom census information was analyzed to determine the ethnic composition of the total population studied. The expected and actual usage of NHS Direct was determined for each section of the population and compared by means of Chi-square analysis and the use of standardized residuals. Females from the white ethnic group used the service more than expected, whereas females from all ethnic groups combined used the service less than predicted. For male callers, Black-African, Indian, Pakistani, Bangladeshi, and Asian groups used the service more than expected. Particularly high usage was observed in Indian and Pakistani groups. The findings of this study show that NHS Direct is being under-used by certain ethnic groups and a difference in usage exists depending on gender. As the NHSD is intended to be the first port-of-call for healthcare advice, determining why certain groups use the service more than others is important. Census data show that the ethnic minority communities have grown significantly in recent years. Information is lacking, however, on the differences in the usage of healthcare services by different ethnic components of the population. A consequence of this situation is that certain decisions relating to health care policy cannot be targeted effectively. This limitation is important as the Race Relations (Amendment) Act 2000 places a statutory duty on NHS organizations to promote race equality, in policy and service delivery. Our findings also raise questions relating to the reliability of some current forms of disease surveillance and also show that NHS Direct data to determine patterns of disease, within the population, will be biased by the uneven usage of the service.