Although serotonergic neural systems have been implicated in the control of anxiety for a number of years, evidence in favour of this role is controversial. The present experiments were designed to further characterize the putative role of serotonin (5-HT) in anxiety, using two pharmacologically validated animal models: the elevated plus-maze and the shock-probe burying tests. If the integrity of 5-HT neural systems is necessary for the expression of 'anxious' behaviors, then disruption of 5-HT systems should produce effects in the plus-maze and shock-probe tests that are similar to those of anxiolytic drugs. In the present experiments, serotonergic function was disrupted in rats, either by chemical depletion using the synthesis inhibitor p-CPA, by inhibitory autoreceptor activation using the selective 5-HT1A receptor ligand 8-OH-DPAT, or by electrolytic lesions of the serotonin-containing, dorsal raphe nucleus. p-CPA and dorsal raphe lesions produced robust anxiolytic effects in the elevated plus-maze and the shock-probe burying tests, whereas 8-OH-DPAT produced anxiolytic effects only in the shock-probe burying test, and 'anxiogenic' effects in the elevated plus-maze test. Although these results generally support the view that serotonin plays a role in the expression of 'anxious' behavior, the opposite effects of 8-OH-DPAT in the two behavioral paradigms suggest that the 5-HT1A receptor subtype exerts differential control over different types of experimental anxiety.