The ''Psychophysiology working group'' of AMDP (Arbeitskreis fur Metho
dik und Dokumentation in der Psychiatrie) intends to establish, a broa
d data base of EEG data in psychiatry. The purpose is to determine whe
ther the EEG can contribute to the diagnostic classification of psychi
atric disorders, to the description of the course of the illness and t
o the prediction of the therapeutic outcome. Since several studies hav
e shown promising results during the last years, members of the workin
g group have investigated literature and written reviews which will be
summarized in this con-tribution. This includes a critical evaluation
of the cited studies. We restrict ourselves to schizophrenia, affecti
ve and anxiety disorder. Out of more than 3000 available contributions
in the world literature 500 have been evaluated by the group, and 216
are taken for this summary. The critical review is the basis for futu
re research of the working group and gives the following working hypot
heses: Under resting conditions anxiety does not show typical deviatio
ns from normality. However, EEG-recordings under experimentally induce
d anxiety do show typical changes: in normal persons there is usually
an increase in desynchronous fast beta-activity (hyperarousal accompan
ied by cognitive coping), whereas patients with anxiety disorder show
hypersynchronized alpha-activity, as it is usually found in deep relax
ation. Thus, hypersynchronized alpha-activity might be an expression o
f a compensatory effort of the system to relax. Anxiolytics induce syn
chronous beta- or sub-alpha activity, which may force the system into
more rigid patterns (relaxation) and thus, could counteract anxiety in
normal persons, if hyperarousal going along with desynchronisation is
not sufficient. Furthermore, benzodiazepines show also an increase in
desynchronous fast beta-activity which could be useful in chronic anx
iety, in order to support cognitive efforts. While depressive disorder
s cause a variety of changes in the wake EEG, that can only partly lin
ked with the clinical picture (neurotic versus endogenous), the sleep
EEG shows a decrease in REM-latency, an increase in REM density and a
decrease in sleep efficiency in patients with depressive symptoms. How
ever, nosological specifity is not established yet. Successful antidep
ressive therapy counteracts these changes. The common feature in the w
ake EEG is the interaction with vigilance dependent variables in both
directions: stimulation and sedation. Schizophrenia shows different pa
tterns for acute and chronic states. In acute schizophrenia there is a
dysrhythmic activity with an increase in beta-activity. This is count
eracted by neuroleptics, inducing slowing of frequency and an increase
of synchronization. In chronic schizophrenics, there is an increased
synchronization with low variability of amplitude. In contrast to the
treatment in acute schizophrenics, in chronic schizophrenia the therap
y response under neuroleptics is indicated by an increase in beta-acti
vity and a decrease in synchronization.