Oral fluid (OF) is being developed as a specimen for the determination of drug intake as an alternative to serum and plasma. It is generally considered as an attractive specimen due to the noninvasive nature of the sampling procedure and the relation to the free fraction of drug in the blood. These features are of particular value in drug treatment of psychiatric disorders. To establish OF for the purpose of monitoring drug therapy, the relationship between concentrations in OF and serum/plasma must be documented. This study explored one promising sampling device and comprised the following 10 drugs: aripiprazole, citalopram, duloxetine, escitalopram, mirtazapine, pipamperone, pregabalin, promethazine, quetiapine, and venlafaxine. For this purpose, 100 paired serum and OF samples were collected from patients undergoing pharmacotherapy and analyzed using a liquid chromatography-tandem mass spectrometry method. A commercial method from Chromsystems for the determination of these drugs in plasma was used and was adapted for OF and ultrafiltrated (Centrifree device) serum. The ratio of each individual pair of samples was used to calculate a mean and SD value between OF and serum free and total concentrations. The OF concentration ratios to serum total fraction differed markedly between substances and differed from 10-fold lower to 8-fold higher. The ratios to serum free fractions were always higher. The relation between the OF and serum concentrations was also evaluated by regression analysis and determination of slopes and correlation coefficients. For all measured relations, there was a statistically significant relation between the OF and serum concentrations. The degree of drug protein binding was in agreement with literature. The aripiprazole, duloxetine, pipamperone, pregabalin, and promethazine concentrations in ultrafiltrated serum were not possible to measure because of low concentrations and nonspecific binding. Despite a strong statistical correlation between OF and serum concentrations observed for most of the studied substances, it is still evident that OF concentrations cannot simply substitute serum/plasma as therapeutic drug monitoring specimen, but rather be considered as a unique specimen. We believe that OF is a promising matrix especially for compliance testing in psychiatry settings. The Greiner Bio-One device used in this study provides a sampling procedure that offers advantages over the available alternatives.