A sequential-sampling plan was developed for making treatment decisions forTrichoplusia ni(Hübner) on cabbage in central Florida. Intensive and scquential samples were taken weekly from randomized replicated plots at different intervals to allow estimation from different population levels. Agreement between intensive and sequential sampling was found in 171 of 178 data sets. A savings of 4 insecticide applications resulted from treatment decisions using the sequential model in a commercial, 2.43– hectare field of cabbage during high looper infestations.