Analysis of bioactive lipids is increasingly useful in clinical studies, and there is a need for non-invasive and easy-to-use sampling methods that meet the demands of reliability. Samples that can be taken by a non-professional and that can be taken repeatedly so as to provide more detailed information about the inflammatory process are often desired. In this study, the feasibility of non-invasive sampling of nasal mucosa and saliva for the analysis of bioactive lipid mediators (e.g. oxylipins and endocannabinoids) was evaluated in a pilot study (n = 10). In a second study, the reliability (relative and absolute) of sampling of these lipid mediators derived from nasal mucosa and from plasma was assessed by calculation of the intraclass correlation coefficient and Bland–Altman’s limit of agreement. Samples were taken at the same time of day on two occasions from a cohort of individuals with and without building-related intolerance (n = 37). Nasal mucosa proved to be a suitable matrix for the analysis of bioactive lipids and was therefore included in the study on reliability together with the plasma samples. Relative reliability varied among the identified oxylipins and endocannabinoids. Arachidonic acid derivatives showed generally better reliability. Absolute reliability measures also varied indicating that only a subset of the oxylipins and endocannabinoids were suitable as biomarkers in either nasal mucosa or plasma and should therefore be used with caution for that purpose.