Abstract In a series of studies, we evaluated the susceptibility of immunoassays for saliva biomarkers to interference effects caused by cotton materials used to absorb saliva during sample collection. Salivary assay results for testosterone, DHEA, progesterone, and estradiol are artificially high, and for sIgA artificially low, when samples are collected using cotton absorbent materials. In contrast, results for salivary cortisol, DHEA–S, and cotinine are not affected by the use of cotton collection methods. The order of individual results from samples collected using cotton versus no-cotton methods for certain markers is not conserved, suggesting that for some biomarkers this collection method can be a significant source of unsystematic error. It was shown, using DHEA as an example, that the cotton interference effect is of sufficient magnitude to attenuate the association between serum and saliva levels. Awareness of this issue is critical to ensure measurement validity in future studies and analyses of archived samples collected using cotton materials.