利用灰色模型ＣＭ（１,１）、三次指数平滑、三次抛物线模型和改进三次抛物线模型四种方法，对五所医院的八组反映门诊和住院工作量的指标进行了建模运算和拟合及外推预测效果的比较分析。最后得出结论认为，对于医院年门诊和住院工作量的预测，当资料分布呈缓慢上升，上升中又有波动趋势时，改进三次抛物线模型优于其它几种方法。 Using the gray model CM (1,1), the cubic exponential smoothing, the cubic parabola model, and the improved cubic parabola model, the modeling, calculation, and fitting of the eight sets of indicators reflecting outpatient and inpatient workload were performed. The comparison of predictive effects. Finally, it is concluded that for the hospital’s annual outpatient and inpatient workload predictions, when the data distribution is slowly rising and there is another fluctuation in the rise, the improved cubic parabola model is superior to the other methods.