Background <p>Adolescent depression is associated with significant morbidity and psychosocial impairment across the lifespan. Robust associations between peripheral inflammatory markers (PIMs) and depression have been found in clinically depressed adults. Studies suggest associations between PIMs and a range of biopsychosocial factors (e.g., stress, sleep) in community samples of adults. No prior studies have examined links between PIMs and a wide array of biological/psychosocial factors (including parent-reported factors, such as parent report of child’s depression severity) in a clinically depressed cohort of adolescents, which might independently inform depression risk and also guide development of interventions to reduce risk and severity of depression.</p> Methods <p>We conducted a cross-sectional study of both depressed adolescents (<i>n</i> = 52) and healthy adolescents (<i>n</i> = 20) and explored associations between 31 biopsychosocial factors and PIMs, specifically tumor necrosis factor alpha (TNFα) and interleukin-6 (IL-6). We also conducted moderation analyses. We also utilized multiverse analyses, examining robustness of results to analytic decisions. Given the described gaps in the field, our overarching scientific objective was to identify biopsychosocial factors that might contribute to variance in associations between PIMs and biopsychosocial measures in depressed adolescents.</p> Results <p>We found that TNFα was reproducibly associated with parent-reported depression severity amongst all adolescents and depressed adolescents. Less reproducible associations between IL-6 and diastolic blood pressure were also found. In moderation analyses, only links between TNFα and parent-reported depression severity were stronger in depressed adolescents.</p> Conclusions <p>We found that specifically in depressed adolescents, TNFα may reflect parent-reported depression severity (which may serve as a more proximal measure of a child’s observable behaviors related to depression). However, few other biopsychosocial variables were linked to PIMs, with small or negligible effect size associations for most examined relationships, suggesting small sample sizes (e.g., n’s &lt; 75) may be insufficient to detect links between PIMs and biopsychosocial variables. Given the use of our multiverse analyses, our analyses can help future researchers focus on understanding potential mechanisms linking PIMs to parent-reported depression severity.</p>