The Ki-67 proliferation index (PI) is a prognostic and surrogate immunohistochemical marker that is used by oncologists to identify the molecular subtypes Luminal A and B in estrogen receptor (ER)-positive breast carcinomas. The 2011 Saint Gallen Expert Consensus Conference recommended that one should adhere to the Ki-67 PI cut-off point of less than 14% for Luminal A subtype and more than 14% for Luminal B subtype. The 2013 novel revision established the KI-67 PI cut-off point of 20% or more for the luminal B subtype. Visual assessment and digital image analysis (DIA) of Ki-67 PI lead to the redistribution of the subtypes in the group of luminal carcinomas, and the same case may be present in different groups according to the method of analysis. Objective — to study the percentage distribution of Luminal A and B subtypes in relation to cases of grade 2—3 invasive unspecified breast carcinoma with ER-positive and HER2-negative receptor status based on the assessment of Ki-67 PI between the visual method and DIA of histological specimens. Material and methods. The investigators selected 64 invasive unspecified breast carcinomas divided into 2 molecular subtypes: luminal A and luminal B according to the immunohistochemistry (ICH-4)-based surrogate classification. Luminol A and luminol B subtypes were separated at the Ki-67 PI cut-off point of 14% and 20, respectively. Ki-67 PI was estimated using DIA, visual assessment, on the basis of the retrospective data of morphopathological conclusion and revision by an independent investigator. Results. DAI revealed that the number of cases of luminal A subtype was 25% higher than those determined by the visual assessment based on the retrospective data of morphopathological conclusion and 7% higher than those through the assessment by an independent researcher using Ki-67 PI cut-off point of 14%. The Ki-67 PI cut-off point of 20% revealed by DAI showed a satisfactory distribution with a 7% difference for retrospective data and 6% Ki-67 PI for an independent investigator. Conclusion. The revealed percentage redistribution of cases of luminal A and B subtypes in invasive unspecified breast carcinomas requires additional analysis with the choice of a single method for Ki-67 PI estimation. The authors propose to consider DAI as one of these methods with the use of digital histology slide scanning and a module implemented in the software for calculation.