[目的]研究香蕉N、P、K肥施用量,为建立香蕉平衡施肥指标体系提供参考依据。[方法]通过对香蕉进行N、P、K肥料“3414”试验,并对结果进行肥料效应回归分析,拟合建立产量三元二次、一元二次和二元二次肥料效应回归模型。[结果]适量的N、P、K肥料对香蕉的增产效应明显,过量施肥造成肥料利用率下降,在生产中N、P、K肥宜选择适宜的用量。建议采用一元二次肥料效应模型计算推荐施肥量,试验土壤的香蕉最佳产量为44.193～45.904t/hm2时,对应的N、P2O5、K2O最佳经济施用量分别为795.1、262.3、1236.9kg/hm2,其N、P、K施肥比例为1∶0.33∶1.55。[结论]试验结果可供与试验地(黏壤土)相类似的土壤类型区参考应用。 [Objective] The research aimed to study the application amount of N, P, K fertilizers in banana and provide reference for establishing the balanced fertilization index system of banana. [Method] The experiment of N, P, K fertilizer “3414 ” on banana was conducted, and the results were analyzed by regression analysis of fertilizer effect. The regression model of yield quadratic, quadratic and binary fertilization was established . [Result] Appropriate amount of N, P and K fertilizers had obvious effect on yield increase of banana. Fertilizer utilization rate decreased with excessive fertilization. Appropriate amount of N, P and K fertilizer should be selected in production. It is suggested that the optimal fertilization rate should be calculated by using the model of one-element secondary fertilizer effect. When the optimum yield of banana in test soil is 44.193-45.904 t / hm2, the optimal economic application rates of N, P2O5 and K2O are 795.1, 262.3 and 1236.9 kg / hm2, the N, P, K fertilization ratio of 1: 0.33: 1.55. [Conclusion] The results of the experiment can be used for reference in the soil type area similar to the test area (clay loam).