BackgroundPreliminary studies report no negative and a possible positive impact of deep brain stimulation (DBS) on cognition of patients with treatment-resistant depression (TRD). However, these studies neither controlled for practice effects nor compared active with sham stimulation.MethodTo address these limitations, we compared 25 TRD patients, who underwent DBS of the ventral anterior limb of the internal capsule (vALIC), with 21 healthy controls (HCs) matched on gender, age and education level. Both groups did subtests of the Cambridge Neuropsychological Test Automated Battery assessing verbal and visuospatial memory, attention, cognitive flexibility, psychomotor functioning, planning and object naming. TRD patients were tested 3 weeks prior to DBS surgery (baseline), 3 weeks following surgery (T1) and following 52 weeks of DBS optimization (T2). HCs were tested at baseline, 6 weeks following baseline (T1) and 20–24 weeks following baseline (T2). Subsequently, TRD patients entered a randomized, double-blind crossover phase, in which they were tested in an active and a sham stimulation phase.ResultsTRD patients did not improve on a test of immediate verbal recognition from baseline to T1, whereas HCs did (group x time: p = 0.001). Both TRD patients and HCs improved over sessions on tests measuring delayed verbal recall, visuospatial memory, planning and object naming (all p < 0.01). Active and sham stimulation did not have an impact on any of the tests differentially.ConclusionsvALIC DBS neither has a lasting positive nor negative impact on cognition in TRD patients. DBS surgery might have a temporary negative effect on verbal memory.