Background: Two examples of clinical research with mistletoe extracts were used to demonstrate essential difficulties in carrying out randomized and placebo-controlled trials. Study 1: In a randomized, placebo-controlled, double-blind study investigating the immunological effects of mistletoe extract, healthy subjects were asked to state whether, in their estimation, they had been treated with verum or a placebo. Due to the intrinsic effects of the mistletoe therapy – local inflammatory reactions at the injection site – 100% of the subjects treated with verum and 77% of those treated with a placebo made a correct assessment of their therapy. Although double-blind trials are preferable from the methodological point of view – above all in QoL research – this study shows that double blinding is barely achievable when the investigated therapy has obvious (side) effects. Study 2: A prospective, randomized, multicenter study of a mistletoe therapy complementary to chemotherapy treatment of breast cancer had to be stopped after a period of 28 months, because it proved impossible to recruit more than 16 patients in six large study centers. With regard to this example and to other failed, GCP-compliant clinical trials on mistletoe therapy we describe which factors interfere with successful clinical trials. One important point, especially in the investigation of complementary cancer treatments, is that cancer patients are unwilling to have their treatment determined by randomization. Many cancer patients in Germany have their own point of view, as to whether a complementary treatment could be of benefit to them or not. Faced with a life-threatening disease they wish to determine this part of their treatment themselves. Conclusion: This background elucidates the need for improving the methodology of non-randomized trials to obtain objective and reliable results even in these fields of clinical research.