The spread of drug-resistant forms of TB dictates the need for surgical treatment in the complex of anti-tuberculosis measures in Russia. Most often, surgical intervention is performed in the case of pulmonary tuberculoma or fibrotic cavitary tuberculosis (FCT). This study is devoted to the search for biomarkers that characterize the course of disease in surgical TB patients. It is assumed that such biomarkers will help the surgeon decide on the timing of the planned operation. A number of serum microRNAs, potential regulators of inflammation and fibrosis in TB, selected on the basis of PCR-Array analysis, were considered as biomarkers. Quantitative real time polymerase chain reaction and receiver operating curves (ROC) were used to verify Array data and to estimate the ability of microRNAs (miRNAs) to discriminate between healthy controls, tuberculoma patients, and FCT patients. The study showed that miR-155, miR-191 and miR-223 were differentially expressed in serum of tuberculoma with “decay” and tuberculoma without “decay” patients. Another combination (miR-26a, miR-191, miR-222 and miR-320) forms a set to differentiate between tuberculoma with “decay” and FCT. Patients with tuberculoma without “decay” diagnosis differ from those with FCT in serum expression of miR-26a, miR-155, miR-191, miR-222 and miR-223. Further investigations are required to evaluate these sets on a larger population so as to set cut-off values that could be applied in laboratory diagnosis.