This study has attempted to validate the Geriatric Depression Scale translated version (Chinese) with a psychiatric outpatient sample (N = 461) of males and females aged 60 or above, from 10 government-maintained psychiatric outpatient clinics between January 1992 and February 1993. Reliabilities and validities were exceptional. Internal consistency reliability was .89 (alpha), and the test-retest reliability was .85 (alpha). Criterion-related (psychiatrist diagnosis) validity was good at .95, and concurrent validity (with CES-D) was .96. Item analysis also confirmed consistency--all 30 items were significantly correlated with the full GDS. However, its sensitivity (70.6%), specificity (70.1%), false negatives (29.4%), and false positives (29.9%), though acceptable, were not as impressive. The overall result has shown that the GDS is generally applicable to the Chinese elderly population and is good for measuring depressive symptoms. The scale can be easily applied in the community by health care professionals. However, further follow-up studies are recommended.