Abstract Background Patients with Crohn’s disease (CD) responding to anti–tumor necrosis factor (anti-TNF) show great variability in serum drug levels, even within the therapeutic range. We aimed at exploring the role of inflammatory, genetic, and bacterial variables in relation to anti-TNF through levels in CD patients. Methods Consecutive CD patients receiving stable doses of infliximab or adalimumab were included. Clinical and analytical parameters were recorded. Cytokine response, bacterial DNA translocation, and several immune-related genes’ genotypes were evaluated, along with serum through anti-TNF drug levels. A linear regression analysis controlled by weight and drug regimen was performed. Results One hundred nineteen patients were initially considered. Five patients on infliximab and 2 on adalimumab showed antidrug antibodies in serum and were excluded. One hundred twelve patients were finally included (62 on infliximab, 50 on adalimumab). Fourteen patients on infliximab and 15 on adalimumab (22.6% vs 30%, P = 0.37) were receiving an intensified drug regimen. C-reactive protein (CRP), fecal calprotectin, Crohn’s Disease Activity Index, leukocyte count, and albumin levels in plasma were not significantly associated with infliximab or adalimumab levels in the multivariate analysis. Serum interleukin-10 (IL-10) levels were directly related to infliximab (Beta = 0.097, P < 0.0001) and adalimumab levels (Beta = 0.069, P = 0.0241). The best multivariate regression model explaining the variability of serum infliximab and adalimumab levels included IL-10. Predicted drug levels by this model robustly fitted with actual drug levels (R2 = 0.841 for infliximab, R2 = 0.733 for adalimumab). Conclusion Serum IL-10 is significantly related to serum anti-TNF levels in CD patients, showing how the disposition of anti-TNF drugs is significantly influenced by the degree of immunological activation.