In order to meet the planning requirements of offshore heavy oil thermal recovery measures, it is necessary to carry out a research on the prediction period of the validity period and decline rate of the steam stimulation.At present, the prediction model of steam stimulation decline rate only considers the influence of reservoir static parameters and steam injection parameters, and does not consider the influence of viscosity coefficient of heavy oil fluid,resulting in the decline rate of steam stimulation of heavy oil fields with different rheological characteristics.The difference in prediction accuracy is large.In view of the large difference in rheological properties of heavy oils with different colloidal and asphaltene contents, the sample set generated by orthogonal design considers the influence of parameters such as viscosity-temperature coefficient, reservoir static and steam injection parameters.Based on multivariate nonlinear regression analysis, a new model for predicting the rate of decline in steam stimulation considering the influence of the viscosity coefficient is established.The application of the example shows that the prediction accuracy of the steam stimulation decline rate of different oilfields by the new model is high, which can meet the needs of the diminishing forecast of steam stimulation in different heavy oil fields.The model provides a fast and accurate prediction method of decline rate for steam stimulation scheme research, effect evaluation and production planning.