A sensitive measure of soil inocuolum potential is needed to evaluate field management of common root rot (Aphanomyces euteiches) in peas (Pisum sativum). A modified rolled-towel (RT) bioassay had been proposed to measure soil inoculum potential in fine-textured soils used for pea production in Minnesota. Homogenized soil was used because organic debris containing the inoculum could not be separated by wet sieving. The poor precision prompted an evaluation of procedures to improve this modified RT bioassay. Seed treatment with a 5% solution of sodium hypochlorite before pea seed germination and plant isolation procedures during the RT bioassay preparation/incubation reduced seedborne contamination and seedling loss to less than 5%. Tests conducted with pasteurized soil that was artificially infested with oospores showed the region of the pea taproot 1 to 2 cm below the seed more susceptible to infection (33% compared with 15% infected seedlings) than the region 1 to 2 cm above the root tip. A soil volume of 1.0 cm3 increased inoculum potential compared with 0.5 cm3 applied to each seedling but did not influence the random error; the 40-seedling compared with the 20-seedling RT bioasssay reduced random error from 18 to 12%. The modified RT bioassay conducted on soil that was artificially infested after steam treatment or without steam treatment showed superior performance when using 40 seedlings compared with 20 seedlings when evaluated for accuracy and precision. Multiple infection theory demonstrated more multiple infections in the RT bioassay with a 0.5 cm3 soil volume applied to each seedling, which shows that soil mass is a factor preventing a higher percentage of infected seedlings. These modifications to the RT bioassay improved the method enough to reduce the random error by one-half compared with using homogenized soil without the proposed modifications.