Several neuroimaging studies have reported 'hypofrontality' in depressed patients performing a cognitive challenge compared to control subjects. Hypofrontality in depression is likely associated with an impaired behavioral performance. It is unclear whether this impaired performance is the consequence or the cause of hypofrontality. Consequently, we proposed to compare the cerebral activity of depressed patients and healthy subjects while controlling for the level of performance. Ten individuals meeting DSM-IV criteria for Major Depression and 10 healthy controls were tested with a verbal version of the n-back task during fMRI scanning. The working memory load was manipulated across the experiment (1,2,3-back) to increase the cognitive demands. fMRI data were acquired on a 1.5-T GE scanner and analyzed using SPM99 software. We did not find any difference between groups in both performance and reaction times for each level of complexity of the n-back task. Depressed patients and control subjects showed bilateral activation of the lateral prefrontal cortex, anterior cingulate and parietal cortex. Activation of these regions was modulated by the complexity of the task. Within this n-back neural network, depressed patients showed greater activation of the lateral prefrontal cortex and the anterior cingulate compared to healthy subjects. This study provides evidence that depressed patients need greater activation within the same neural network to maintain a similar level of performance as controls during a working memory task. Our findings suggest that depression may impair the cognitive capacity of depressed patients by recruiting more brain resources than controls during cognitive control.