There is accumulating evidence that benzodiazepines (BZD)--agents widely used as anxiolytics and hypnotics-could be regarded as "natural" drugs since they have been found in trace amounts also in plants, various tissues of different animal species and even humans. The biosynthesis of such BZD is still unknown and the hypothesis is favoured that they may be of plant origin. Besides diazepam (D) and its major metabolite desmethyldiazepam (DD) several other BZD (e.g. delorazepam, deschloro-diazepam, delormetazepam, isodiazepam, lormetazepam, oxazepam) could be detected. In some cases identification of these compounds was accomplished by specific mass spectrometry (GC-MS) and for quantification various methods have been applied resulting in different concentrations which range for D from about 0.005 to 1 ng/g and for DD from 0.01 to 0.5 ng/g. It is very unlikely that these trace amounts exert any direct pharmacological effects and at the moment only speculations upon their physiological/biological role are possible. Recently BZD-receptor binding activity equivalent to surprisingly high levels of more than 900 ng/ml was found in cerebrospinal fluid of patients with advanced hepatic encephalopathy. As long as the structure of this binding activity has not been elucidated no firm conclusions can be drawn from these findings. If pertinent analytical problems (e.g. drug-free biological material; exact quantification by internal standard techniques) are solved and if the source(s) of BZD are established it might be possible to answer also the critical question whether "endogenous" or "natural" BZD play any (in-) direct role in the regulation of CNS activity.