Abstract-Two types of conditioned behaviors were investigated for the purpose of evaluating anxiolytic drugs. A conditioning procedure used was an active avoidance in poorly-performing mice. Chlordiazepoxide, diazepam, chlorazepate and meprobamate increased the avoidance rate, while chlorpromazine, haloperidol and nortriptyline did not produce such an effect. The effect of diazepam was potentiated by r-aminobutyric acid (GABA) and aminoxyacetic acid (AOAA) and antagonized by picrotoxin and thiosemicarbazide, but was influenced little by spiroperidol, a-methyltyrosine, phenoxy benzamine and levallorphan.In addition, the effect of other anxiolytics was potentiated by AOAA and antagonized by picrotoxin.Biperiden, methamphetamine, caffeine and morphine also induced an avoidance enhancement, which was not influenced by AOAA. A drug discrimination experiment was also performed using a milk-reinforced two-lever operant method.In the rats trained to discriminate phenobarbital from saline, diazepam produced a dose-related phenobarbital-lever selection, which was potentiated by AOAA and antagonized by picrotoxin.Chlordiazepoxide, chlorazepate and meprobamate also elicited responses on the phenobarbital-lever.On the other hand, haloperidol, nortriptyline, biperiden, methamphetamine, caffeine and morphine produced a saline-lever selection, at the doses tested. These results suggest that, among several drugs tested, the avoidance enhancement and discriminative response control by anxiolytics may be closely linked with the GABA system. Among the numerous methods used to evaluate anxiolytic drugs, the conditioning procedures such as conditioned suppression (1, 2) and Geller's conflict test (3, 4) are available for indexes of conditioned "anxiety" and to demonstrate marked difference between anxiolytics and the other classes of drugs. It is most difficult, however, in these tests to produce appropriate response patterns, because environmental contingencies such as re inforcement schedule, duration and frequency of discriminative stimuli and the extent of punishment determined the degree of response suppression and drug effects. Under some conditions, amphetamine (5, 6), morphine (7,8), chlorpromazine (9, 10) and reserpine (11) also alleviate the response suppression induced by punishment.The effects of anxiolytics on the other conditioned behaviors have also been examined by several investigators. Takaori et al. (12,13) reported that benzodiazepine anxiolytics enhanced the conditioned avoidance response of poorly-performing animals in both Sidman avoidance and shuttle box avoidance procedures. However, this effect was not specific for anxiolytics since several non-anxiolytics were also effective in similar avoidance tests