Behavioural models of depressive illness must attempt to satisfy at least the four following criteria: similarity of inducing conditions, similarity of the behavioural State induced, similarity of the underlying neurobiological mechanisms and similarity of clinically effective treatments. The fourth criterion is probably the most relevant because fairly reliable information about the efficacy of treatments, particularly drugs is already available. Thus judicious testing of known drugs can serve as an important tool for analysing the behavioural changes observed in different behavioural models. The accompanying paper illustrates this point by describing pharmacological analysis of separation and social dominance phenomena in monkeys and of acute and chronic stress effects in rats.
