Although the mechanism of action of current antidepressant drugs is well known, 30% of patients remain refractory to treatment. There are examples of recent failures of antidepressant clinical development programs that reinitiate the discussion on the reliability, or predictability of animal models of major depressive disorder. The comorbid expression of depression is well known in many neurological and psychiatric diseases, including drug abuse. Symptoms such as anhedonia, hypo- or hyperlocomotor activity, sleep disturbances, and weight loss can be measured in animals, but the overreliance on such readouts may lead to misinterpretations of their relevance to the clinical situation. Existing models are mainly based, or could be considered to have an overreliance, on the putative involvement of monoaminergic systems. Data recorded from these existing models do result in clinical candidate nomination, but when such compounds reach the clinic, data often do not fully meet expectations. From a preclinical point of view, the integration of the range of established and novel technologies, such as monitoring dynamic changes in extracellular neurotransmitters, behavioral readouts, electrophysiological recordings, and brain scanning (e.g. using functional magnetic resonance imaging and spectroscopy, PET, SPECT imaging) is critically needed in the development of new animal models, as is the translation of such approaches to the clinic and vice versa. This integration is needed across the gamut of indication areas of key interest.