The hypothesis linking decreased serotonin transmission to reduced anxiety as the mechanism in the anxiolytic activity of benzodiazepines conflicts with most clinical observations. Serotonin antagonists show no marked capacity to alleviate anxiety. On the other hand, clinical signs of reduced serotonergic transmission (low 5-HIAA levels in the cerebrospinal fluid) are frequently associated with aggressiveness, suicide attempts and increased anxiety. This brief review attempts to reconcile these human and animal findings by investigating whether anxiety reduction or increased impulsivity are likely to account for animal behavioral changes associated with decreased serotonergic transmission. The effects of manipulating central serotonin on experimental anxiety paradigms in animals (punishment, novelty) are reviewed and compared to the effects of anti-anxiety drugs. Anxiety seems neither necessary nor sufficient to induce control by serotonergic neurons on behavior. Further evidence suggests that behavioral effects of anxiolytics thought to be mediated by decreases in anxiety are not caused by the ability of these drugs to reduce serotonin transmission. Blockade of serotonin transmission, especially at the level of the substantia nigra, results in a shift of behavior towards facilitation of responding. This behavioral shift is particularly marked when there is competition between acting and restraining response tendencies and when obstacles prevent the immediate attainment of an anticipated reward. It is proposed that serotonergic neurons are not only involved in behavioral arousal but also in enabling the organism to arrange or tolerate delay before acting. Decreases in serotonin transmission seem to be associated with the increased performance of behaviors which are usually suppressed though not necessarily because of the alleviation of anxiety which might contribute to the suppression.