Abstract: Three key problems are encountered when applying the quantitative analysis technology of the whole petroleum system to evaluate oil-gas resource by genetic method. One is how to input the thickness, total organic carbon (TOC) and hydrogen index (HI) of representative source rocks and the corresponding hydrocarbon generation kinetics model during the grid calculation. Another is how to establish and select reasonable parameters of hydrocarbon generation kinetics model of the source rock. The third is how to reasonably analyze the efficiency of hydrocarbon expulsion, migration and accumulation. Based on analyzing the data of source rock as well as oil and gas in petroliferous basins, in combination with the evolution and development of hydrocarbon generation and expulsion algorithm, this paper discusses how to promote the solution of the above three problems. Firstly, the representation of input parameters in the grid calculation of hydrocarbon generation is solved by the hierarchical evaluation of organic facies of source rock. Secondly, the hydrocarbon generation kinetics model is calibrated from the actual big data of source rock and the simulated experimental data. Thirdly, the efficiency of hydrocarbon expulsion, migration and accumulation is determined from the large data of residual hydrocarbon and migration loss, so as to reduce the uncertainty of hydrocarbon resource evaluation by genetic method.