Affective disorders such as major depression are among the most prevalent and costly diseases of the central nervous system, but the underlying mechanisms are still poorly understood. In recent years, it has become evident that alterations of the stress hormone system, in particular dysfunctions (hyper- or hypo-activity) of the hypothalamic-pituitary-adrenal (HPA) axis, play a prominent role in the development of major depressive disorders. Therefore, we aimed to generate a new animal model comprising these neuroendocrine core symptoms in order to unravel parameters underlying increased or decreased stress reactivity. Starting from a population of outbred mice (parental generation: 100 males and 100 females of the CD-1 strain), two breeding lines were established according to the outcome of a 'stress reactivity test' (SRT), consisting of a 15-min restraint period and tail blood samplings immediately before and after exposure to the stressor. Mice showing a very high or a very low secretion of corticosterone in the SRT, i.e. animals expressing a hyper- or a hypo-reactivity of the HPA axis, were selected for the 'high reactivity' (HR) and the 'low reactivity' (LR) breeding line, respectively. Additionally, a third breeding line was established consisting of animals with an 'intermediate reactivity' (IR) in the SRT. Already in the first generation, i.e. animals derived from breeding pairs selected from the parental generation, significant differences in the reactivity of the HPA axis between HR, IR, and LR mice were observed. Moreover, these differences were found across all subsequent generations and could be increased by selective breeding, which indicates a genetic basis of the respective phenotype. Repeated testing of individuals in the SRT furthermore proved that the observed differences in stress responsiveness are present already early in life and can be regarded as a robust genetic predisposition. Tests investigating the animal's emotionality including anxiety-related behavior, exploratory drive, locomotor activity, and depression-like behavior point to phenotypic similarities with behavioral changes observed in depressive patients. In general, HR males and females were 'hyperactive' in some behavioral paradigms, resembling symptoms of restlessness and agitation often seen in melancholic depression. LR mice, on the other hand, showed more passive-aggressive coping styles, corresponding to signs of retardation and retreat observed in atypical depression. Several morphometric and neuroendocrine findings further support this view. For example, monitoring the circadian rhythm of glucocorticoid secretion revealed clearly increased trough levels in HR mice, resulting in a flattened diurnal rhythm, again adding to the neuroendocrine similarities to patients suffering from melancholic depression. Taken together, our results suggest that distinct mechanisms influencing the function and regulation of the HPA axis are involved in the respective behavioral and neurobiological endophenotypes. Thus, the generated HR/IR/LR mouse lines can be a valuable model to elucidate molecular genetic, neuroendocrine, and behavioral parameters associated with altered stress reactivity, thereby improving our understanding of affective disorders, presumably including the symptomatology and pathophysiology of specific subtypes of major depression.