Questions of research ethics always arise when planning a correspondence test to study discrimination in the market place. However, the issue is addressed relatively little in published correspondence tests with authors usually referring to the two seminal articles written in this field (i.e. Banton (1997) and Riach and Rich (2004)). Since then correspondence testing has become more widespread and the technique is increasingly relying on the internet to find and send applications. It is therefore necessary to revisit the question of ethics in correspondence testing. This paper addresses the ethical issues that researchers are facing in correspondence tests that study discrimination in hiring decisions in the labour market in particular. It provides a short overview on the development of research ethic guidelines. The main part of the paper focuses on the ethical issues that arise in correspondence testing, looking at questions of covert research, potential problems (regarding voluntary participation, informed consent, deception, entrapment of employers, employer’s rights), possible solutions and technical challenges. Looking at specific country examples, decisions by ethical commissions and national legal frameworks are considered. These show that testing has to be renegotiated depending on the national context, and, in the case of Germany, legal implications of correspondence testing are discussed. The paper concludes that correspondence testing, if planned carefully and executed responsibly, does not violate research ethics in social sciences.