The study aimed to analyze hsa-miR-16-5p, hsa-miR-143-3p, hsa-miR-423-5p, hsa-miR-137-3p, hsa-miR-489-5p, hsa-miR-520-3p, hsa-miR-486-5p, and hsa-miR-200a-3p expression in the serum of patients with invasive non-functioning pituitary adenomas (NFPAs) and prolactinomas, as candidates for non-invasive biomarkers. The study included 62 patients with NFPAs and 18 with macroprolactinoma who qualified for transsphenoidal surgical resection. MicroRNAs were isolated from serum samples. The expression levels of hsa-miR-16-5p, hsa-miR-143-3p, hsa-miR-423-5p, hsa-miR-137-3p, hsa-miR-489-5p, hsa-miR-520-3p, hsa-miR-486-5p, and hsa-miR-200a-3p were determined using TaqMan MicroRNA assays. The statistical analyses were performed with MedCalc. The total concentration of microRNA was significantly lower in NFPAs than in the CG (p = 0.0419). ROC curve analysis showed that the cutoff point of miRNA lower than 10.73 predicted the PA (sensitivity = 70.0%; specificity = 57.7%; AUC = 0.629; p = 0.052). No correlation between selected miRNAs and tumor type was found: hsa-miR-143-3p (p = 0.4610), hsa-miR-16-5p (p = 0.8767), and hsa-miR-423-5p (p = 0.1459). miRNA expression also did not correlate with invasiveness (cavernous or sphenoid sinus invasion, optic chiasm compression). Although the total expression of microRNA was significantly lower in NFPAs, hsa-miR-16-5p, hsa-miR-143-3p, and hsa-miR-423-5p are not useful as non-invasive biomarkers in patients with invasive non-functioning pituitary adenomas and prolactinomas.