Background. In the last decades, the incidence of breast cancer (BCa) in young women has been increasing steadily. The quantitative indicators of expression of collagen, which play important role in stromal microenvironment, and their association with the age and survival rates of BCa patients have not been yet definitively clarified. Aim. To investigate the relationship between the COL1A1 gene expression at the mRNA and protein levels in BCa tissue and the clinicopatological features and survival rates of BCa patients of different age groups. Materials and Methods. The study was conducted on the clinical material of 50 patients with stage I—III BCa. COL1A1 gene expression at the mRNA and protein levels in BCa tissue were studied using the real-time PCR and immunohistochemical methods, as well as the bioinformatic analysis (UALCAN and Kaplan — Meier Plotter databases). Results. The bioinformatic analysis showed that BCa tissue is characterized by 6.0 times (p < 0.05) higher level of COL1A1 mRNA compared to normal breast tissue. The correlation of COL1A1 expression at the mRNA and protein levels with the molecular subtype of neoplasms was demonstrated. According to Kaplan — Meier Plotter database, a low level of expression of COL1A1 protein level in BCa tissue is associated with lower rates of relapse-free survival of patients. The ex vivo study of the clinical material revealed a decrease in COL1A1 protein expression in tumor tissue of young patients with BCa of T3 category (p < 0.0374), low differentiation grade (p < 0.0163) and basal molecular subtype (p < 0.0001). A correlation between the expression of COL1A1 at the mRNA and protein levels and the expression status of estrogen receptors (p < 0.0001) and progesterone receptors (p < 0.0040) was established. The relapse-free 3-year survival rate of young BCa patients is significantly lower in the presence of a low COL1A1 optical density index in the tumor tissue. Conclusions. The identified relationship between COL1A1 expression and such indicators of BCa malignancy as tumor size, differentiation grade, molecular subtype, receptor status, and the recurrencefree survival of patients indicates the prospects of its use to predict the aggressiveness of the BCa course in young patients.