Widespread invasive annual grasses, cheatgrass (Bromus tectorum L.) and Japanese brome (Bromus japonicus Thunb.), fluctuate greatly in abundance and compete with native species. Fire and herbicide have each provided various levels of short-term control. We tested the individual and combined effects of fall fire and the herbicide aminopyralid on annual brome to determine whether combined treatment increased or extended control. Treatments were a factorial arrangement of two fire (no fire; fall fire with 2-yr return interval) and three herbicide (no herbicide; alternate-yr herbicide; annual herbicide) treatments with five replications. Across years, fire doubled bare ground and reduced litter cover to half of that with no fire. Fire had no effect on germination of brome seed produced after fire. Fire reduced brome biomass one or two growing seasons after fire. The first 3 study yr, brome biomass was < 29 kg·ha^–1 with fire and was 132 kg·ha^–1 without fire. During the final year, brome biomass increased and was similar (704 kg·ha^–1) across fire treatments. Nonbrome biomass was 19% greater with fire during 2015 and 2018. Nonburned plots shifted from dominance by C₃ perennial grass to dominance by bromes by the last year. C₃ perennial grass maintained dominance through 2017 with fire and was codominant with bromes during 2018. Aminopyralid reduced brome germination each year it was applied but did not affect brome biomass. C₃ perennial grass dominated in both herbicide treatments through 2017. During 2018, bromes dominated with no herbicide or alternate-yr herbicide and were codominant with C₃ perennial grass with annual herbicide treatment. All treatment combinations reduced forbs compared with nonburned, no-herbicide treatment. Long-term control of annual bromes requires long-term commitment to repeated treatment. The combination of fall fire and spring application of aminopyralid did not extend annual brome control under the study conditions.