Qualitative research methods are specifically designed to handle complexity, in particular the complexity of human behavior. I argue that empirical research in software engineering has reached a state of maturity in which we are beginning to address truly complex problems. The complexity of these new problems lies not only in a new interest in the human role in software development, but also in the fact that there are still plenty of difficult technical issues to be addressed. This presentation gives some examples of how qualitative methods can usefully be incorporated into empirical studies of software engineering, and the advantages of doing so.