In making pest management decisions 2 types of sampling errors are: (1) the decision to treat when no treatment is necessary and (2) the decision not to treat when treatment is needed. A type 1 error results in a waste in the cost of chemicals, their application, etc. while a type 2 error may result in losses in yield and profits. Sequential sampling plans have been developed which set both errors at 80% allowing decisions to be made in an average of 15 minutes per field. The confidence level could be set higher but more time would be required, so a compromise has been reached between the 80% level of confidence and expensive sampling time. Improved prediction of crop loss may be obtained by using both the number of pests that will cause unacceptable losses (the action level) and the number of natural enemies required to regulate pest abundance (the inaction level). Use of these levels with sequential sampling eliminates much guesswork and adds confidence to investment decisions.