T he prevalence of depression following stroke varies between 14% and 19%, as compared to 10% in the general population. 1 Compelling reasons exist for effective treatment of depression in relationship to stroke. These include the fact that depression is a risk factor for stroke, and that post-stroke depression (PSD) is associated with higher mortality. 2,3 Antidepressant drugs have been shown repeatedly to result in better outcomes compared to placebo in PSD, including the critical finding of lowered mortality. 4,5 Early studies of tricyclic drugs demonstrated efficacy, but side effects and poor tolerance limited the use of these compounds. Such concerns have been lessened by newer generation drugs, although they are not without their shortcomings. For these and other reasons, many depressed people self-medicate with complementary and alternative types of treatment, among which herbal products are popular. 6 While there is a fair amount of information on the efficacy of St. John's wort (Hypericum perforatum), less is known about other herbal products for depression. The report by Li and associates in this issue (pp. 841-846) is a welcome addition and follows earlier positive studies of Free and Easy Wanderer Plus (FEWP) in bipolar and unipolar depression. 7 The authors report unambiguous advantages for FEWP and the selective serotonin reuptake inhibitor (SSRI) fluoxetine over placebo in PSD. In addition, FEWP acted more rapidly than fluoxetine and, at endpoint, was superior to fluoxetine on daily activities, an effect that has been inconsistently seen in the PSD literature. 8 Fifty percent of all placebo-controlled antidepressant trials in PSD have been conducted in Asia, primarily in China, and these contributions have advanced the field considerably. 8 This study of FEWP by Li and associates suggests a new and potentially helpful approach. Among its strengths was the inclusion of an active control, thereby establishing that the study population was indeed "antidepressant responsive." If this had been a two-arm comparison without placebo and there was no difference in outcome, we would be left to guess if FEWP was ineffective or if the sample was not preferentially drug responsive and therefore inappropriate to test the drug. The large, adequately powered sample was also impressive, making it one of the largest trials of its kind in PSD. The very high retention rate is also unusual. With only 4 premature dropouts (3%), this study stands apart from most others in this population, where attrition rates are typically between 10% and 35% among trials of conventional antidepressants. 8 In a prior trial of FEWP in non-stroke related depression, the dropout rate was over 30%. 7 In explaining their anomalous retention rate, the authors suggest the role played by massage as a means of attenuating nausea, which can certainly affect patient compliance. This creative approach in managing nausea deserves greater attention. Li and colleagues also suggest that the study sample might have been unusually motivated, although in ...