Failure io obtain convincing results in psychiatric genetics may partly be
attributed to the fact that in molecular biology and genetic epidemiology h
as not been followed by art equivalent developement in phenotypic descripti
on. Instead or relying entirely on classical nosographical approaches,we ar
gue that identifying more homogeneous forms if diseases through a candidate
symptom approach among affected subjects and subclinical trails among non-
affected relatives (endophenotype approach) may yield better results. Examp
les where these strategies have already been fruitful in other complex dise
ases are presented. Focusing on vulnerability traits might stimulate the re
definition of traditional psychiatric syndromes and should help to bridge t
he the gap between clinical and experimental approaches.