Artificial insemination (AI), as an efficient assisted reproduction technology, can help the livestock industry to improve livestock and poultry breeds, optimize production performance and improve reproductive efficiency. AI technology has been widely used in pig production in China, but boar fertility affects the effectiveness of AI, and more and more studies have shown that there are significant differences in the fertility of boars with similar semen quality indicators. Therefore, this study aimed to identify biomarker molecules that indicate the level of boar fertility, which is important for improving the efficiency of AI. In this study, we collected 40 mL of ejaculates per boar used for extracellular vesicle (EV) characterization in 20 boars and identified 53 differentially expressed miRNAs by small RNA sequencing, of which 44 miRNAs were up-regulated in the high-fertility seminal EVs compared with low-fertility seminal EVs, and nine miRNAs were down-regulated. miR-26a was most significantly down-regulated in the high-fertility group compared to the low-fertility group, and it was hypothesized that this miRNA could be used as a biomolecular marker of semen reproductive performance. To further determine the effect of miR-26a on sperm function, we successfully established a miR-26a overexpression model and found that miR-26a reduced sperm viability, motility, acrosome integrity, plasma membrane integrity and ATP levels. Bioinformatics analysis and dual luciferase reporter analysis revealed that miR-26a directly targets High mobility group A1 (HMGA1). In conclusion, miR-26a can be used as a biomarker to identify high and low fertility in boar semen.