1 The acute behavioural effects of the a2-adrenoceptor antagonists, yohimbine, idazoxan and delequamine (RS-15385-197) were compared in two tests of exploratory behaviour in the rat, operated in tandem. These were the elevated X-maze test (5 min) and a modified holeboard test (12 min), which comprised a holeboard arena with a small roof in one corner as a 'refuge'. Rats were first placed into this corner, thus enabling measurements of initial emergence latency and the number of forays. The experiments were always done with a concomitant vehicle control group, with 10-12 rats per group, and with the treatment blinded. 2 In order to validate the tests, the effects of representatives of four classes of psychoactive agents were examined, viz. picrotoxin (anxiogenic), chlordiazepoxide (anxiolytic), (+)-amphetamine (stimulant) and diphenhydramine (sedative). The modified holeboard tended to be more sensitive than the measurement of total arm entries in the elevated X-maze at detecting drug effects on exploratory behaviour, but unlike the X-maze it could not clearly identify each class of agent. Thus, picrotoxin (5 mg kg-', i.p.) reduced total arm entries and open arm exploration in the X-maze (P<0.02) and suppressed most measures of activity in the holeboard (P<0.05); chlordiazepoxide (7.5 mg kg-1, i.p) increased total arm entries and open arm exploration (P<0.02) in the X-maze, without clear-cut effects in the holeboard; (+)-amphetamine (1 mg kg-, i.p.) had no significant effects in the X-maze, but increased most holeboard activities (P<0.05), and diphenhydramine (30 mg kg-', i.p.) reduced total arm entries in the X-maze (P<0.002) and hole exploration in the holeboard (P<0.05). 3 The actions of yohimbine most closely resembled those of picrotoxin. In the elevated X-maze, yohimbine (3 mg kg , i.p.) decreased the total number of arm entries (P<0.02); a larger dose (10 mg kg-', i.p.) also reduced time spent on the open arms (P<0.02). In contrast, delequamine (3 mg kg-, i.p.) and idazoxan (3 mg kg-', i.p.) had no effect. 4 In the partially-shaded holeboard, yohimbine (3 mg kg-', i.p.) suppressed hole exploration (P<0.05); a higher dose (10 mg kg-', i.p.) increased emergence latency (P<0.002) and virtually abolished all activity. Delequamine (3 mg kg-, i.p.) and idazoxan (3 mg kg-', i.p.) did not influence emergence latency or holeboard activities.5 The extent of the blockade of central a2-adrenoceptors achieved during the tests was assessed by the ability of the doses used to reverse mydriasis induced by clonidine (300 Ag kg-', s.c.) in anaesthetized rats. At a dose of 3 mg kg-, i.p., delequamine and idazoxan produced a rapid, sustained reversal of the clonidine response (by 87 ±2 and 86 ± 2% respectively, 30 min after injection) whereas yohimbine produced a partial reversal of only 43 ± 13%. The higher dose of yohimbine used in the exploratory tests (10 mg kg'-, i.p.) was required in order to achieve 77 ± 4% reversal of clonidine-induced mydriasis.6 We therefore conclude that blockade of central a2-adrenoceptors per se...