Venlafaxine, a serotonin and norepinephrine reuptake inhibitor (SNRI), produces significantly higher remission rates in depressed patients than do the selective serotonin reuptake inhibitors (SSRIs). In this analysis of pooled data, we explored the relationship between differences in treatment efficacy, early improvement of symptoms, and severity of baseline anxiety in depressed patients treated with either venlafaxine or fluoxetine. A pooled analysis was performed on data from 1,454 outpatients with major depression from five double-blind, randomized studies comparing the 6-week efficacy of venlafaxine (542 patients) with fluoxetine (555 patients). The Hamilton rating scale for depression (HAM-D) total and item scores were analyzed at different treatment times up to 6 weeks. Venlafaxine and fluoxetine both produced statistically significant higher response and remission rates compared with placebo starting from week 2 for response and weeks 3 to 4 for remission. Venlafaxine was statistically significantly superior to fluoxetine from week 3 until week 6 in respect of response rate, and from week 2 until week 6 for remission rate. After 1 week of treatment, greater improvement in individual symptoms was observed in the depressed mood, suicide, and psychic anxiety items of the HAM-D scale for both venlafaxine- and fluoxetine-treated patients compared with placebo. Improvement in psychic anxiety was statistically significantly greater with venlafaxine than with fluoxetine. The presence of baseline psychic anxiety correlated significantly to treatment outcome when analyzing the remission rates. In depressed patients with moderate anxiety (HAM-D psychic anxiety score < or = 2), venlafaxine statistically significantly increased remission rates compared with placebo from week 4 until week 6, while a significant effect of fluoxetine on remission rates was observed starting at week 6. Remission rates in the severely anxious depressed patients (score > 2) were statistically significantly higher with venlafaxine than placebo starting from week 3 until the end of the study period, but no difference could be observed between fluoxetine and placebo. Baseline severity of psychic anxiety had a significant impact on remission rates after treatment of patients diagnosed with depression. Venlafaxine's superior remission rates in the more severely anxious patients and its ability to improve psychic anxiety as early as week 1 compared with fluoxetine suggest that venlafaxine's early efficacy on anxiety symptoms may be the basis for its superior efficacy in depression.