Preferences include the choices made by individuals when presented with options for treatment for depression, and the system of beliefs and views that underlies those choices. They are informed by the experience of previous treatment by individuals, their family and friends, information from medical professionals and the media, and incorporates biases and ideologies present within the population. Although the randomised controlled trial is generally considered to be the optimal method for evaluating the effectiveness of health care interventions, [1] patients may become less motivated to follow the treatment protocol if they are not allocated to their preferred treatment. Consequently, the relevant arms of the study may appear less effective as a result. Further, following an invitation to join a clinical trial, patients may refuse randomisation and be excluded from the trial if they have strong treatment preferences, leading to the introduction of bias and restricted ability to generalise the results, as participants may not be representative. Considerable demand has been shown by patients for psychological treatments for the treatment of depression in primary care. However, two recent studies have not demonstrated a relationship between being allowed to choose treatment and short-term depression outcome. These two studies explored primary care patients treated with antidepressants or counselling, and non-directive counselling, cognitive-behaviour therapy or usual general practitioner care. Further work is needed to determine the effects of preferences within different study designs and to explore the views of both professionals and patients using appropriate qualitative designs.