The neural basis of mood disorders is still incompletely understood. Syndromes classified as mood disorders are composed of broad and quite inhomogeneous combinations of a multitude of psychological and physical signs andsymptoms, which typically vary across time, are modified by comorbid conditions, personality traits, sociocultural background and genetic factors. While many effective therapy options exist, still a large proportion of patients with depression and related disorders do not sufficiently respond to treatment. In addition, there is a lack of information that could help to clearly identify patient-related factors optimising individual treatments. The emergence of brain imaging technologies during the past decades is leading to a rapidly growing data-base providing sophisticated information about the neuroanatomy of mood disorders. The currently available information from neuroimaging studies in mood disorders starts to converge into some generalisable patterns of structural and functional brain changes. Although some inconsistencies exist between cross-sectional neuroimaging studies, an increased prevalence of white matter and periventricular hyperintensities appears to be associated with mood disorders. In addition, neuroimaging studies suggest structural and functional alterations in prefrontal cortex and cingulate cortex, the amygdala, hippocampus and other portions within the limbic system, and in the basal ganglia.These areas receive rich serotonergic, noradrenergic and dopaminergic projections and play a central role in the regulation of the hypothalamic-pituitary adrenocortical system. Both are fundamental to current conceptions of antidepressant drug action. Some of these cross-sectional changes appear to be associated with severity and/or duration of the illness, and with outcome during treatment. For example, antidepressants appear to have a protective effect on hippocampal volume loss in major depression, a finding that relates to the capacity of antidepressants to normalise the hypothalamic-pituitary adrenocortical system and to reduce the level of corticosteroids, which in turn can be toxic to hippocampal neurons. It is unquestioned that neuroimaging approaches represent a promising lead in understanding the neural basis of mood disorders. Future directions of brain imaging research in such complex diseases, however, should attempt to more directly associate structural, resting state metabolic and task-dependent functional changes, both in cross-sectional and longitudinal investigations. Nevertheless, the increasing understanding of brain circuits that are specifically associated with certain aspects of the disease provides particularly useful information for translational research, which will foster novel hypotheses that are mutually testable in clinical and preclinical settings. Notably, the available evidence summarised in this article is based on group studies, which do not allow to draw inferences for individual patients. One of the long-term perspectives is that patterns of specific dysfunctional brain activity may be used to select and optimise individually tailor-made treatments. The non-invasiveness and increasing technical sophistication of brain imaging techniques and the synergistic combination of them, keeping in mind the many caveats also briefly outlined in this article, and securing a careful control of clinical and other possible confounds, hold great promise for this challenge.