One of the «low-cost» methods of the welltests is the analysis of production and pressure in unsteady filtration modes. After interpreting the data on production and pressures in mechanized wells with the availability of information about the pressure at the pump intake, an assessment of the current reservoir pressure and an analysis of the causes of production changes for individual low-permeability reservoirs can be performed. For the implementation of the analysis of production and pressure, information on the initial reservoir pressure is required; therefore, the search for new methods for estimating the initial reservoir pressure is an urgent task. The paper proposes to combine the hydrodynamic study of a well with a stepwise change in operating modes with the stage of putting a mechanized well into operation. This combination will reduce costs and significantly reduce production losses associated with conducting «traditional» welltests with a well shutdown. However, the implementation of a study with a stepwise change of modes when putting a well into operation may lead to a decrease in the reliability of determining reservoir parameters. The paper discusses an approach to determining the scenario for conducting such a study. The approach is based on the generalization of multivariate calculations of various scenarios for putting wells into operation in the hydrodynamic simulator «RN-KIM». Based on the results of the analysis and generalization of data, matrices for evaluating research scenarios have been constructed, according to which it is possible to choose an acceptable option for implementing a study with an assessment of reservoir pressure with a certain reliability. The results of the work also make it possible to assess the reliability of reservoir pressure from wells that have already been removed, in which field dynamic data on pressures, debits, etc. has been recorded. An approach is proposed for replicating matrices for estimating the reliability of the initial reservoir pressure without additional multivariate calculations and using previously calculated matrices.