Therapeutic drug monitoring (TDM) is extremely helpful in individualizing dosage regimen of drugs with narrow therapeutic ranges. It may also be beneficial in the case of drugs characterized by serious side effects and marked interpatient pharmacokinetic variability observed with leflunomide and its biologically active metabolite, teriflunomide. One of the most popular matrices used for TDM is blood. A more readily accessible body fluid is saliva, which can be collected in a much safer way comparing to blood. This makes it especially advantageous alternative to blood during life-threatening SARS-CoV-2 pandemic. However, drug’s saliva concentration is not always a good representation of its blood concentration. The aim of this study was to verify whether saliva can be used in TDM of teriflunomide. We also developed and validated the first reliable and robust LC-MS/MS method for quantification of teriflunomide in saliva. Additionally, the effect of salivary flow and swab absorptive material from the collector device on teriflunomide concentration in saliva was evaluated. Good linear correlation was obtained between the concentration of teriflunomide in plasma and resting saliva (p < 0.000016, r = 0.88), and even better between plasma and the stimulated saliva concentrations (p < 0.000001, r = 0.95) confirming the effectiveness of this non-invasive method of teriflunomide’s TDM. The analyzed validation criteria were fulfilled. No significant influence of salivary flow (p = 0.198) or type of swab in the Salivette device on saliva’s teriflunomide concentration was detected. However, to reduce variability the use of stimulated saliva and synthetic swabs is advised.