The relevance of the research is caused by the need to create methods for operational processing the results of non-stationary researches in horizontal wells with hydraulic fractures, and increasing accuracy of parameter estimates of oil reservoirs and decreasing shut-in time. The main aim of the research is to develop the models and algorithms for identifying filtration flow regimes, defining filtration characteristics and reservoir pressure in the processing of non-stationary researches in horizontal wells with hydraulic fractures. The methods are based on the researches in the field of hydrodynamic studies of wells with hydraulic fractures system analysis, system modeling taking into account additional information and expert estimates, function optimization, linear algebra. Using integral system of wellbore pressure models with time-dependent variables, with account of additional prior information and expert estimates about reservoir permeability and reservoir pressure, the problem of filtration flow identification was solved. Testing of models and algorithms for flow identification and determination of reservoir and fracture parameters was carried out on the basis of field data from non-stationary researches of horizontal oil wells with hydraulic fractures by pressure buildup curves using the software package Saphir. The results. Processing non-stationary research results by pressure buildup curves of horizontal wells shows that the proposed models and algorithms for identification of filtration flow regimes allow determining the effective length of horizontal wellbore, lateral permeability, reservoir pressure and skin factor in the process of research, under conditions of partial or complete absence of late radial flow on pressure buildup curves, significantly reducing shut-in time.