The prevalence of major depression in stage 5 chronic kidney disease (CKD) varies between 14 and 30%. Patients with CKD who are depressed have a worse quality of life, are hospitalized more often and die sooner than those who are not depressed. Antidepressant drugs are effective in the general population, but whether they improve outcomes in CKD is uncertain. Drug pharmacokinetics are altered in CKD, which may necessitate dose adjustment. We aimed to systematically review available evidence of the pharmacokinetics, efficacy and safety of antidepressant drugs when used in patients with CKD3 to CKD5 (CKD3-5). This is a systematic review of randomized clinical trials and observational studies examining antidepressants in patients with CKD3-5, regardless of whether or not patients are on dialysis. Through comprehensive searches of seven databases, we identified all studies examining pharmacokinetic properties or clinical outcomes in patients with CKD3-5. One author assessed studies for eligibility and quality and extracted all data. Antidepressant drugs were the studied intervention. The main outcomes were pharmacokinetic parameters, clinical outcomes such as response to treatment, reduction in depression severity and adverse events. We identified 28 studies evaluating pharmacokinetic parameters in CKD for 24 antidepressants. Sparse and heterogeneous data precluded informative meta-analysis. Drug clearance in CKD3-5 was markedly reduced for selegiline, amitriptylinoxide, venlafaxine, desvenlafaxine, milnacipran, bupropion, reboxetine and tianeptine. We identified one randomized controlled trial (RCT) in 14 patients on haemodialysis for fluoxetine versus placebo which showed no difference for efficacy and safety measures. One other RCT of escitalopram versus placebo in 62 patients on haemodialysis provided no efficacy data. There were nine non-randomized trials, all suggesting benefit for the antidepressant under investigation. Side-effects were common, but mild in most patients. The limitations of this review include the scarcity of randomized trial data, the small size of the observational studies and possibility of publication bias. In addition, study selection and data extraction were done by one reviewer only, increasing the risk for errors made in handling of the data. Dose reduction in CKD3-5 is necessary for selegiline, amitriptylinoxide, venlafaxine, desvenlafaxine, milnacipran, bupropion, reboxetine and tianeptine. The evidence on effectiveness of antidepressants versus placebo in patients with CKD3-5, and with the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV)-defined depression is insufficient, and in view of the high prevalence, a well-designed RCT is greatly needed.