Six potato fields were randomly sampled using quadrat areas ranging from 10 m2 to 4,000 m2. A composite soil sample consisting of 20 soil cores was taken from each of 20 quadrats of each unit area. All samples were assayed for Verticillium dahliae and nematodes. The area in which a composite sample was taken had little effect on estimated mean and variance of V. dahliae until quadrat size was ≥1,000 m2. Then variances dramatically increased with size. With Pratylenchus spp., there was no systematic relationship between quadrat size and variance. Seven additional fields were divided into 2,000 or 4,000 m2 grids, and each grid was sampled in both spring and fall for V. dahliae and plant-parasitic nematodes (Pratylenchus spp. and Meloidogyne hapla). Density of V. dahliae was lower, and in some cases, degree of aggregation was higher, in the spring than in the fall. Sampling for V. dahliae in the fall required 2 to 7 composite samples to estimate mean density within 25% of the true mean, while in the spring, 10 to 19 samples were required. Assays of Pratylenchus spp. required a similar number of samples in spring and fall surveys, while M. hapla required fewer samples with spring than with fall sampling when estimating densities >20 second-stage juveniles per 100 cm3 of soil.