This study aimed to prove the validity of a mixture of chemicals, including salts, small organic molecules, mucin, and α-amylase, as saliva surrogate (“artificial saliva”) for assessing leakage of methacrylate monomers and other constituents from dental materials. To achieve this, we developed and validated a liquid chromatography-tandem mass spectrometry (LC-MS/MS) method for the quantification of 2-hydroxyethyl methacrylate (HEMA), triethylene glycol dimethacrylate (TEGDMA), diurethane dimethacrylate (UDMA), bisphenol A glycerolate dimethacrylate (BisGMA), diphenyl(2,4,6-trimethylbenzoyl)phosphine oxide (TPO), bisphenol A (BPA), and five homologues of ethoxylated bisphenol A dimethacrylate (BisEMA EO2–6) in unstimulated and artificial saliva, and compared their concentrations in the two saliva media following either spiking with a mixture of the compounds or incubation of test specimens of printed biomaterials. Test specimens were immersed in unstimulated/artificial saliva, incubated at 37 °C for 24 h, and saliva aliquots were extracted with methanol and subsequently analyzed by LC-MS/MS. The method was validated with regard to matrix effects, linearity, selectivity, lower limits of quantification (LLOQ), precision, bias and combined measurement uncertainty (u′). The performance characteristics of the method were comparable for unstimulated and artificial saliva samples. The combined u’ for individual chemicals at a concentration of 10 × LLOQ were within the range of 5.3–14 % for unstimulated saliva and 6.9–16 % for artificial saliva, except for the BisEMA homologues. Combined u’ for the latter were 27–74 % in unstimulated saliva, and 27–79 % in artificial saliva. There was no detectable release of BPA from the test specimens, and the TPO concentrations were mainly below the LLOQ. TEGDMA and UDMA were detected in the highest quantities, and at comparable concentrations in the unstimulated and artificial saliva. For all BisEMA homologues, the release was higher in unstimulated saliva than in artificial saliva. The study showed that the artificial saliva model can be a suitable replacement for native saliva, but might underestimate leakage of more lipophilic methacrylates.