Sequential sampling plans based on binomial distribution are presented for determining the need for measures to control tarnished plant bug,Lygus lineolaris(Palisot de Beauvois), in strawberry fields. Because no single theoretical distribution could describe data counts in every field, a double plan was developed by fitting the data to a regression equation. The equation related the mean number of nymphs per blossom cluster to the proportion of infested blossom clusters in the field. Four density classes were recognized: high (above the action threshold level of 0.25 nymphs per blossom cluster), medium, low, and very low. The sequential sampling plan with an action threshold of 0.25 nymphs per blossom cluster was compared with randomly tapping 100 blossom clusters and calculating an average count of nymphs per blossom cluster. In 100 different fields, the two techniques gave similar conclusions on 88% of occasions.