The exotic grass-fire cycle is degrading semiarid rangelands, such as the vast areas of shrub-steppe in North America now invaded by fire-promoting cheatgrass. Chemical- or bio-herbicides are sprayed onto soils to inhibit the invaders, but information on chemical- or bio-herbicide impacts to soil microbial communities is limited. We asked how the soil-microbiome responded to the bioherbicide Pseudomonas fluorescens strain ACK55 in comparison to the separate and combined effects of a conventional pre-emergent chemical herbicide, imazapic, in two cheatgrass-invaded sagebrush-steppe sites. First-year microbial responses were evaluated using targeted sequencing of the 16S and LSU rRNA genes for bacteria+archaea and fungi, respectively, and were related to plant-community responses. A strong cheatgrass reduction with imazapic at one site was accompanied by a small shift in bacteria+archaea (16S) community composition with no effect on microbial alpha diversity, and this shift was small in comparison to natural microbiome variation between sites. ACK55 was not detected in soil a year after application, and it caused only transient and marginally significant reductions in annual grass cover accompanied by small reductions in soil fungi species richness. Full-length sequencing of the ACK55 16S rRNA gene and phylogenetic analyses revealed that ACK55 is more likely P. salmonii than P. fluorescens. Knowledge gaps remain on the duration and consequences of microbial-community shifts with imazapic and why molecular analyses showed ACK55 did not persist in soils. Confusion regarding microbial biopesticides can result where isolation, effectiveness testing, commercial release, and regulation are not guided by molecular taxonomic analyses.