The relationship between ethnicity and health is attracting increasing attention in international health research. Different measures are used to operationalise the concept of ethnicity. Presently, self-definition of ethnicity seems to gain favour. In contrast, in the Netherlands, the use of country of birth criteria have been widely accepted as a basis for the identification of ethnic groups. In this paper, we will discuss its advantages as well as its limitations and the solutions to these limitations from the Dutch perspective with a special focus on survey studies. The country of birth indicator has the advantage of being objective and stable, allowing for comparisons over time and between studies. Inclusion of parental country of birth provides an additional advantage for identifying the second-generation ethnic groups. The main criticisms of this indicator seem to refer to its validity. The basis for this criticism is, firstly, the argument that people who are born in the same country might have a different ethnic background. In the Dutch context, this limitation can be addressed by the employment of additional indicators such as geographical origin, language, and self-identified ethnic group. Secondly, the country of birth classification has been criticised for not covering all dimensions of ethnicity, such as culture and ethnic identity. We demonstrate in this paper how this criticism can be addressed by the use of additional indicators. In conclusion, in the Dutch context, country of birth can be considered a useful indicator for ethnicity if complemented with additional indicators to, first, compensate for the drawbacks in certain conditions, and second, shed light on the mechanisms underlying the association between ethnicity and health.