It has been suggested that antidepressants that increase noradrenergic transmission should be better than serotonergic antidepressants at treating fatigue in depression. We conducted a secondary analysis of an RCT in which patients with depression were randomly assigned to reboxetine (a noradrenaline reuptake inhibitor [NARI]) or citalopram (a selective serotonin reuptake inhibitor [SSRI]). We investigated the difference between citalopram and reboxetine in treating the symptom of fatigue in depression, and also in treating depression with high levels of baseline fatigue. We found no difference between citalopram and reboxetine in terms of improvement in fatigue at six weeks (0.11, 95% confidence interval (-0.28, 0.49); p = 0.59:), or at 12 weeks. Using the Beck Depression Inventory at 12 weeks as the outcome measure, we found some evidence in support of our hypothesis that reboxetine is more effective than citalopram in treating depression in those with high baseline fatigue (interaction term: -2.87, 95% confidence interval (-5.15, -0.60); p = 0.01). We conclude that there is no evidence of any difference between reboxetine and citalopram in their efficacy in treating fatigue as a symptom of depression, but that reboxetine may be more effective in treating depression with high levels of fatigue. Fatigue might be useful in the prediction of response to NARIs or SSRIs.