The nervous system is rich in miRNAs, indicating an important role of these molecules in regulating processes associated with cognition, memory, and others. Therefore, qualitative and quantitative imbalances involving such miRNAs may be involved in dementia contexts, including Late-Onset Alzheimer's Disease (LOAD). To test the viability of circulating miRNAs (c-miRNAs) as biomarkers for LOAD, we proceed accordingly to the following reasoning. The first stage was to discover and identify profile of c-miRNAs by RNA sequencing (RNA-Seq). For this purpose, blood serum samples were used from LOAD patients (n = 5) and cognitively healthy elderly control group (CTRL_CH) (n = 5), all over 70 years old. We identified seven c-miRNAs differentially expressed (p ≤ 0.05) in the serum of LOAD patients compared to CTRL_CH (miR-10a-5p; miR-29b-2-5p; miR-125a-5p; miR-342-3p, miR-708-5p, miR-380-5p and miR-340-3p). Of these, five (p ≤ 0.01) were selected for in silico analysis (miR-10a-5p; miR-29b-2-5p; miR-125a-5p; miR-342-3p, miR-708-5p), for which 44 relevant target genes were found regulated by these c-miRNAs and related to LOAD. Through the analysis of these target genes in databases, it was possible to observe that they have functions related to the development and progress of LOAD, directly or indirectly connecting the different Alzheimer's pathways. Thus, this work found five promising serum c-miRNAs as options for biomarkers contributing to LOAD diagnosis. Our study shows the complex network between these molecules and LOAD, supporting the relevance of studies using c-miRNAs in dementia contexts.