Based on the agri-ecosystem analysis

the early-warning index system and risk degree evaluation for calamity of agriculture pest were constructed

which consisted of the related factors including pests

natural enemies

cropping system

climate

social and economical factors

level of integrated pest and so on. According to the expression of risk in the calamity process

these factors arc classified into three layers

namely risk resource

risk sign and risk case

respectively. Each factor of risk could be measured by grades with the methods of fixed quantity or qualitative analysis. The weight coefficient of each factor should be given according to the degree related to final calamity. The models calculating the integrated early warning index and risk grade were proposed. They could be used to help optimizing decision for integrated pest control and decreasing the degree of calamity. Cotton bollworm （Helicoverpa armigera Hubner） is one kind of important pests which could cause serious calamity. As an application example

this paper introduced the technique of early-warning for calamity of cotton bollworm in detail. Based on the fundamentals of system analysis and by applying the structural analysis/design and object - oriented programming method

a practical computer system （MLCYJJC-CDROM） of the early-warning for cotton bollworm calamity and the auxiliary control decision ?making with program language VB6. 0 was developed. It is easy to operate

understand

extend and modify. In this system

the subsystem of the early-warning for cotton bollworm calamity consists of the early warning index setup and risk degree evaluation for calamity of cotton bollworm. The subsystem of the auxiliary control decision-making could not only get the information from the early-warning subsystem

but also contains a lot of information to help auxiliary control decision -making. The application of this system was discussed.