Abstract: Using conventional logging data to accurately calculate the permeability of drill-stem test (DST) can greatly improve the accuracy of productivity predictions of offshore heterogeneous sandstone reservoirs. Based on this, the influence of sedimentary diagenesis and pore structure of Huizhou Sag on reservoir permeability were comprehensively considered from macroscopic and microscopic perspectives, respectively. In this work, a logging interpretation model of absolute permeability was built for different reservoir types. Forward analysis results show that reservoirs with different permeability contrast in perforated intervals have significantly different contributions to productivity.The synthetic logging permeability was calculated by weighted summation of permeability at different levels of the reservoir, and the weight coefficient was constrained to highlight the contribution of favorable reservoirs to productivity. An iterative analysis was performed with a differential evolution algorithm to yield the optimal solution of the equation. This method has been applied in 72 oil layers in Huizhou Sag for productivity prediction. The productivity from 48 layers was found greater than 100 m³/d, and the proportion of layers whose relative prediction errors within 30% was 90%. In addition, 24 layers had the productivity of 10–100 m³/d, among which the layers whose relative error was less than 50% accounted for 79% of the oil layers. This study indicates that the productivity prediction method based on permeability synthesis technology can guide the decision-making of offshore field tests and operations to reduce the exploration cost.