An attempt is made to relate drug and non-drug antidepressive interventions to brain processes. In the present context two concepts are proposed: vulnerability towards depressogenic factors and depression as a state of the brain. Accordingly, it is assumed that the current antidepressants make the brain less vulnerable by inhibiting monoaminergic neurons and by reducing the number or efficacy of their postsynaptic receptors. In contrast, electroconvulsive therapie and sleep-deprivation affect primarily the state of the brain and can produce spectacular (but often short lasting) improvements. It is proposed that the combination of the two aspects may lead to fast and lasting antidepressive effects.