Background: Treatment for depressive elders is a challenge due to the complexity in clinical manifestation, medical comorbidity and etiologies. There has been scant of the literature regarding treatment outcomes and their correlates among the elderly patients with major depressive disorder at acute stage in Taiwan. We conducted a preliminary 12-week open-labeled trial of antidepressants for the late-life major depressive disorder, and examined the predictors of patients with poor outcome. Methods: Sixty elderly patients with late-life major depressive disorder were recruited for a 12-week open-labeled study of acute antidepressants treatment. Depressive severities were assessed using 17-item Hamilton Depression Rating Scale for Depression (HAMD17) at weeks, 0, 2, 4, 8, and 12-week. Treatment remission were defined by HAMD17 ≦7 and Clinical Global Inventory (CGI) ≦2. Antidepressants were prescribed according to clinical judgment. The intent-to-treat population using the last observation carried forward (LOCF) method was used for analysis. Results: Among the patients achieving remission during the trial, the proportions of remitted patients of each visit were: 15.4% at the 2-week, 23.1% at the 4-week, 50% at the 8-week, and 11.5% at the 12-week. We found that the elderly patients with late-onset of first major depressive episode, higher scores of mood factor, and sleep factor were likely to be non-remitted from antidepressant treatments. Conclusions: In this study, only 40% of elder patients could complete 12-week acute antidepressant treatments. Near half of all study participants reached remission. In summary, late-onset of the disease, severe depressed mood, and sleep disturbances could be the predictors for non-remission. But the study results need to be validated in future studies with larger sample sizes in the numbers of the study subjects as well as group comparation and randomization in design.