To compare the efficacy and safety of augmenting paroxetine with risperidone, buspirone, valproate, trazodone, or thyroid hormone in patients with treatment-resistant depression (TRD), 225 patients with retrospectively and/or prospectively identified stage II TRD were randomly assigned to receive an 8-week treatment of paroxetine 20 mg/d augmented with risperidone 2 mg/d (n = 45), sodium valproate 600 mg/d (n = 39), buspirone 30 mg/d (n = 46), trazodone 100 mg/d (n = 47), or thyroid hormone 80 mg/d (n = 48). The primary outcome was the remission rate defined as the 17-item Hamilton Rating Scale for Depression score of 7 or less at the end of study. Secondary outcomes included remission rate based on the Self-rating Depression Scale score of 50 or less at the end of study, response rate based on 17-item Hamilton Rating Scale for Depression total score of 50% improvement or greater from baseline, and the change in scores of Clinical Global Impression-Improvement scale, the Short Form 36 Health Survey, and the Life Satisfaction Rating Scale. The remission rates were 26.7% for risperidone, 48.7% for valproate, 32.6% for buspirone, 42.6% for trazodone, and 37.5% for thyroid hormone. There was no statistical significance among treatment arms in remission rates, secondary outcome measures, and adverse events. Risperidone, valproate, buspirone, trazodone, or thyroid hormone augmentation to paroxetine 20 mg/d was effective and well tolerated in Chinese patients with TRD. Large-sample studies are warranted to support or refute these findings.