The integration of oxytocin (OT) into behavioral science seems to hold considerable promise for advancing our understanding of human health and development but methodological issues restrict the measurement of OT in large studies, in everyday social settings, or when repeated sampling is required. Measuring OT in saliva could overcome many of these limitations. In this paper, we rigorously evaluate the feasibility of doing so. A series of experiments leads to the conclusion that saliva does not contain oxytocin in measurable amounts, and that OT is not a valid salivary biomarker when measured by currently available immunological methods. Levels of immuno-reactive OT in saliva are primarily due to non-specific interference with antibody-antigen binding. We can state with a high degree of certainty that measurement of OT in saliva does not yield meaningful indices of individual differences or intra-individual change.