Ethics in engineering and science has become a hot topic not only on the agendas of academic institutions and funding agencies, but also among scientists and engineers themselves and the general public. Analysis of misconduct cases shows that fundamental issues concerning proper methodology as well as ethics are at stake. Traditionally, questions of methodology and ethics have been treated more or less as separate issues, or as being related but fundamentally different, while practitioners tend implicitly to see ethics as the underpinning of their methodology. I contend that methodological and ethical normativity are linked in a very fundamental way. I argue that the relationship between method and ethics deserves to be addressed more explicitly and can form the basis of a new approach towards ethical issues in science and engineering. In particular, a virtue-ethical based approach, defining the figure of the 'good practitioner', has many advantages. As an example, I consider the issue of the function and utility of ethical codes of conduct. Practice shows that these codes are seldom known to practitioners and, that when they are, practitioners have not internalised them. In other words, they do not match with notions of proper behaviour as experienced by scientists and engineers themselves. I argue that as long as the ethical codes are top-down regulations of an institution, rather than the 'living morality' of the virtuous community of scientific practice, they will remain of limited applicability.