There is a large amount of literature on the optimization of gas-water alternating injection parameters using a single scale, such as core or reservoir, and the problem is that the research results are limited to a single scale and the optimization results of different scholars for the same injection parameter are quite different. For this reason, we use indoor core physical simulation and reservoir numerical simulation techniques to synthesize core, well profile and reservoir scales, and discuss the differences of research results in different scales on the basis of research results in the three scales. The results show that the research results of the three scales are consistent in order of magnitude, but there is a certain gap in the specific values, and the distribution range of the research results of the three scales is basically the same if they are evaluated according to the distribution intervals for the preferential evaluation; from the point of view of inhibiting the gas scampering, the injection parameters of CO2 non-mixed-phase substitution can be appropriately conservative, for example, the gas injection rate can be further reduced by 20% to 30% on top of the results of the optimization of the cores and sections; the injection parameters suitable for the research work area can be reduced by 20% to 30%. 30%; the final optimization results of CO2 non-mixed-phase gas-water alternating injection parameters suitable for the study area are 1:2 gas-water ratio, 0.002PV gas section plug length, 0.12PV total gas injection volume, and 6.8 m3/d gas injection rate; the multi-scale optimization process adopted in this study can provide a reference for the optimization of the parameters of non-mixed-phase gas-water alternating injections in similar tight oil reservoirs.