Postmortem blood samples may not accurately reflect antemortem drug concentrations, as the levels of some drugs increase due to postmortem redistribution (PMR). The brain has been suggested as an alternative sampling site. The anatomically secluded site of the brain limits redistribution and prolongs the detection window, thereby enabling sampling from deceased individuals where blood is no longer suitable for analysis. We report concentrations in brain tissue and blood from 91 cases for the four antidepressants citalopram, duloxetine, mirtazapine and sertraline. The cases were classified according to their role in the cause of death, as follows: (A) concentrations where the drug was the sole cause of fatal intoxication; (B) concentrations where the drug contributed to a fatal outcome; and (C) concentrations where the drug was not related to the cause of death. The analytical method was successfully validated in brain tissue in terms of linearity, process efficiency, precision and accuracy. Quantification of analytes was performed by ultra-performance liquid chromatography with tandem mass spectrometry. Correlations between blood and brain concentrations were achieved with R2-values between 0.67 and 0.91. The following median brain-blood ratios were obtained: 3.71 for citalopram (range: 1.4-5.9), 11.0 for duloxetine (range: 5.0-21.6), 1.53 for mirtazapine (range: 1.02-4.71) and 7.38 for sertraline (range: 3.2-14.2). The S/R ratio of racemic citalopram was the same in brain (0.80) and blood (0.85), whereas the median citalopram/N-desmethylcitalopram ratio was higher in brain (9.1) than blood (4.1). The results of this study may serve as reference concentrations in brain for forensic cases.