Counts of com plants infested by European com borer, Ostrinia nubilalis (Hubner), fall army worm, Spodoptera frugiperda (J. E. Smith), and com earworm, Helicoverpa zea (Boddie), were described by binomial and beta-binomial probability distribution functions. Two sequential classification sampling plans were constructed using thresholds of 15 and 5% infested plants. Stop lines for these sampling plans were constructed using a binomial model. The precision with which these plans classified the proportion of infested plants when sample counts were distributed according to a beta-binomial distribution was determined using simulation. When counts were beta-binomially distributed, precision was reduced. Sampling plans were judged to be acceptable though because the overall operating characteristic function obtained after repeatedly sampling a population 2 or 3 times was shifted to the left on the proportion of infested plants scale resulting in an increased likelihood of reaching a decision to intervene. Field evaluation of the sampling plans revealed that they classified pest infestation very precisely and on average, resulted in a 45% savings in sample costs when compared with a fixed sample size.