The behavioral effects of the benzodiazepine (BZP)-receptor partial inverse agonists, beta-CCE and FG 7142, are reviewed and the claim that these compounds possess "anxiogenic" properties is examined. Results obtained from human studies and global observations in animals, as well as those from experiments on aggression in animals or from studies of pentylenetetrazole discrimination cannot be considered conclusive. Contradictory findings have been obtained in studies using animal testing procedures derived from BZP-sensitive models of anxiety and in newer experimental situations and these are discussed from various theoretical perspectives: (1) the ability of the models to measure increased anxiety; (2) the possible ability of the drugs to reveal latent anxiety which generalizes from a punished to an otherwise non-fearful component of a testing procedure ("spreading anxiety"); (3) anxiety produced by a pro- or pre-convulsant state. Finally, several hypotheses are considered to account for the behavioral effects of beta-CCE and FG 7142 without assuming anxiogenic properties. These include the possible existence of different forms of anxiety, rate dependency, and drug-induced motivational changes. It is concluded that available data are insufficient to strongly support the notion that FG 7142 and beta-CCE are the anxiogenic drugs "par excellence" they are often claimed to be.