When used in a pest control decision-making context, protocols for obtaining sample information on insect numbers can be evaluated via three properties: the operating characteristic (OC), the average sample number (ASN), and the availability of methods for calculating the OC and ASN. On the average, sequential sample plans have lower ASN's, and OC's are comparable to fixed sample-size procedures. The two most commonly used sequential procedures for insect sampling are Wald's probability ratio test and a test proposed by Iwao. Each of these methods has limitations. An alternative to sequential procedures is double sampling. Double sample plans developed for the Colorado potato beetle,Leptinotarsa decemlineata(Say), on Long Island, N.Y., have OC's comparable to a modified form of Iwao's sequential method and fixed sample size methods. They also have ASN's that at times are lower than those of the sequential method.