Abstract Background Saliva represents an attractive biospecimen as it is pain-free, simple for at home collections, and patients can provide multiple samples over time to provide time resolved data. However, few diagnostic tests have been developed using saliva due to poor stability and comparatively low concentration of analytes. Stabilization of biomarkers at ambient temperature in saliva would enable labs to take advantage of the benefits of saliva sample collections, while minimizing the risks posed by the degradation of already scarce analytes. Methods Matched samples of neat serum, saliva, and saliva collected with Saliva Multi-omic Device (SMD) preservative were collected from 30 volunteers (15 male & 15 female) ages 18–64 with no known morbidities. The Olink Inflammation panel was used to examine 92 inflammation-related proteins. The method utilized proximity extension assay technology, wherein antibody probe pairs reacted with their respective target proteins to generate a polymerase chain reaction target sequence through proximity-dependent DNA polymerization. This sequence was measured using real-time polymerase chain reaction and normalized in two steps, producing a semi-quantitative Normalized Protein eXpression (NPX) unit. To evaluate analyte stability specimens were measured initially and then again after 24 h at room temperature (23 ± 2°C). Results Protein expression (NPX) units for 36 of the 92 inflammatory markers had higher concentrations in saliva compared to serum. Serum (78/92) and saliva 78% (72/92) showed similar coverage for the number of markers in the panel where at least 75% of subjects had values exceeding the limit of detection (LoD). SMD preservation buffer improved the average stability of inflammatory protein biomarkers by 19.6% compared to neat saliva and 10.2% compared to serum over 24-hours at room temperature. However, the relative stability of individual between protein markers [DV1] [HM2] varied. For proteins in SMD buffer, initial concentrations of analytes were highly predictive [HM3] of values 24 h later, having Pearson correlation coefficients > 0.7, for 23 of the analytes and greater than 0.5 for 47. The change in protein marker in serum over time was only moderately predictive of the protein’s stability in saliva with or without preservative, having correlations of 0.33 ± 0.20 and 0.47 ± 0.19 (mean ± standard deviation), respectively. Conclusion This study demonstrates potential utility of saliva in proteomic applications utilizing the Olink platform. The SMD preservation buffer stabilized inflammatory biomarkers in saliva samples likely through protease inhibition and antimicrobial action. Preservation of self-collection salivary samples by subjects without cold storage enables noninvasive sampling and reliable measurements even when samples are collected by users at remote locations or under challenging conditions.