When having a continuous free choice in their home cages between water and alcohol- or drug-containing drinking solutions, rats first develop a controlled consumption of the psychotropic compound and, after several months, lose their control over drug taking. After a long period of abstinence, they reveal an excessive, compulsive drug intake. Adulteration of the drug-containing solutions reduces the doses taken by controlled consumers, but not those of the excessive drinkers, they can therefore be regarded as addicted. These animals show a pre-intake motor restlessness that may be related to craving. In two studies with putative anti-craving agents (the dopamine D2 receptor agonist lisuride and the D2 receptor antagonist flupentixol) we treated alcohol-addicted and non-addicted rats and observed the effects on alcohol taking, alcohol seeking and brain neurotransmission. These two investigations paralleled clinical studies, in both cases the results could be predicted correctly ("pro-craving" effect of both pharmaceutics). Differences between "symptomatic" and possible "causal" therapies are discussed, approaches towards a causal therapy according to an "imprinting"-model of an addition are suggested.