According to the breeding behavior of bollworm, its anniversary activi ty can be divided into four phases: the Pupa in autumn, overwintering, eclosion in spring and damage in summer. For each phase, by selecting some me teorological parameters which is of entomological and statistical meaning a mult ieffectiveness normalized meteorological prediction model of bollworm and anal ysisevaluation indices are developed by normalized processing and weighted com bination of parameters. It shows that occurrence of bollworm is closely related to the meteorological conditions. The complex correlation coefficients of each phase are 0.778-0.878, with the fit of 96.7%. The model has a good result i n the practice.It concludes that the nondimensional proc essing of meteorological parameters is not only favorable to its combination ea ch other, but also to increase the precise of model prediction. However the bio logical meanings of these combination parameters are to be defined through biolo gical experiments.