<p>MicroRNAs (miRNA) are considered promising biomarkers for inflammatory and neoplastic diseases in fecal samples, but despite reproducible miRNA detection, there is no consensus on appropriate normalization in fecal samples. In this study, we aimed to explore an endogenous normalizer for miRNA analysis in fecal samples. For this purpose, we performed a multistep study with 3 independent cohorts. In the first step, we investigated miRNA profiling using serial dilutions of blood and fecal samples from healthy subjects. In the second part, we tested the analysis using a cohort of IBD patients (<i>n</i> = 30) with active disease and in remission and 59 patients with different liver pathologies or healthy subjects. Finally, the data were validated using an independent cohort of patients with liver disease (<i>n</i> = 360). After extraction, miRNAs were analyzed using Affymetrix GeneChip microarray technology in the screening and confirmation cohorts, and TaqMan qPCR was performed for validation. Analysis of the blood stool mixture cohort revealed 5 miRNAs that were stable in all three subgroups. Among them, miR-638 was found to have a Log₂FC of 0.0004 and was selected for further analysis as one of the best studied miRNAs from previous reports. The stability of miR-638 in microarray analysis was demonstrated in two independent cohorts of patients with IBD and liver disease. We observed no effect of age or other factors on miR-638 levels in fecal samples. Furthermore, its relatively high concentration in feces and its stability against potential blood contamination may offer great advantages over the previously used miR-16, suggesting miR-638 as a potential endogenous fecal normalizer.</p>