The validity of the Beck Depression Inventory (BDI) was assessed in a
group of 150 neurological inpatients using Receiver Operating Characte
ristic analysis and DSM-III-R as external criterion. As regards depres
sive disorders as a whole, it was found that the best trade-off betwee
n sensitivity and specificity was the cutoff score of 20. The discrimi
nating ability of the BDI for major depressive disorder was quite sati
sfactory at the cutoff score of 29 contrary to the dysthymic disorder
in which the discriminating power of the BDI was not acceptable. In co
nclusion, the use of the BDI in neurological settings is useful with c
utoff scores depending on the research purposes.