Abstract The test sites consisted of low-income housing units located in Indianapolis, Ind. Treatments were randomly assigned to each apartment before inspection. Evaluation areas within the test site consisted of the kitchen in general, cabinetry, refrigerator, and utility area. Infestations of each area were estimated by researchers using the visual count method. The number of cockroaches seen generally around the room was counted, then each possible harborage area was closely examined. The count for each site was considered to be the sum number of cockroaches counted throughout the site at the time of pretreatment. Test materials were applied to areas where pretreatment counts were made. All applications were applied as a normal cockroach cleanout procedure would demand (i.e., cracks and crevices, baseboards, beneath and behind sinks, stoves and refrigerators, around plumbing intersections of shelves and walls). Quantitative information in flushing activity was collected by visual counts of cockroaches in the general kitchen vicinity 5 and 30 min after treatment. Each treatment was replicated 4 times.