Inoculum density of C. crotalariae was determined by elutriation of samples from 288 quadrats (three 96-sample replicates) in a 723-m2 peanut field in Bladen County, NC North Carolina [USA] in June 1980. Five frequency distributions (Poisson, negative binomial, Thomas double Poisson, Neyman type A, and Poisson with zeroes) were tested for goodness-of-fit to frequency class data for inoculum density. The negative binomial distribution model best described the frequency class data. Values of the k parameter, an index of aggregation, were 2.21, 2.77, and 2.09 for the 3 replicate samples and 2.31 for the combined 288 samples, which indicated a clumping or clustering of inoculum in soil. Nine simulated soil sampling methods, differing in area covered and shape of path, were analyzed for efficiency in estimating C. crotalariae populations from the field. Two methods utilizing diagonal paths (sample size, 16 or 32) had sample means within 5% of the population means. Random samples did not give accurate population estimates.