A 3-year study (2000-2002) in southeastern Wisconsin was conducted to assess the effects of Bacillus sphaericus applied for mosquito control on nontarget wetland invertebrates. The experimental design consisted of control and treatment sites (that were applied by helicopter with Vectolex CG), each in 2 vegetation habitat types: reed canary grass marsh (Phalaris arundinacea) and cattail marsh (Typha spp.). In each of these areas, a predetermined number of timed (30-sec) D-frame aquatic net samples containing vegetation, detritus, and invertebrates were collected 1 day before spraying and 72 h after spraying to detect for effects. We examined and compared 5 bioassessment measures to determine if there was an effect of B. sphaericus on nontarget organisms during each of the sampling years. The metrics tested were (1) mean taxa richness (the mean number of all taxa), (2) mean diversity (combines taxa richness and abundances in a summary statistic; i.e., Shannon Index [H'I]), (3) Diptera richness (minus mosquitoes) as a proportion of all other taxa richness (Diptera/others richness), (4) Diptera abundance (minus mosquitoes) as a proportion of all other invertebrate abundance (Diptera/others abundance), and (5) functional group changes in percent collector-gatherers, collector-filterers, scrapers, shredders, and predators. When Vectolex was applied during 6 treatments at the labeled dosage rate in the above habitats in Brookfield, WI, no detrimental effects to nontarget organisms could be attributed to this microbial insecticide. Much of the variation in the control vs. treatment and pre vs. post plots was attributed to factors other than the effects of B. sphaericus on nontarget organisms, such as the time of sampling, natural variation that occurs in such diverse habitats as canary grass and cattail marshes, and water depth, which varied among years.