Wildfires in the western US increasingly threaten infrastructure, air quality, and public health. Prescribed (“Rx”) fire is often proposed to mitigate future wildfires, but treatments remain limited, and few studies quantify their effectiveness on recent major wildfires. We investigate the effects of Rx fire treatments on subsequent burn severity across western US ecoregions and particulate matter (PM2.5) emissions in California. Using high-resolution (30-m) satellite imagery, land management records, and fire emissions data, we employ a quasi-experimental design to compare Rx fire-treated areas with adjacent untreated areas to estimate the impacts of recent Rx fires (Fall 2018-Spring 2020) on the extreme 2020 wildfire season. We find that within 2020 wildfire burn areas where Rx fires were used prior to 2020, burn severity changed by -16% (p < 0.001) and smoke PM2.5 emissions changed by -101 kg per acre (p < 0.1). Rx fires in the wildland-urban interface (“WUI”) were less effective in reducing burn severity and smoke PM2.5 emissions than those outside the WUI. Overall, Rx fires led to a net reduction of -14% in PM2.5 emissions, including those from the Rx fires themselves. The proposed policy of treating one million acres annually in California could reduce smoke emissions by 655,000 tons over the next 5 years, equivalent to 52% of the emissions from 2020 wildfires. Our analysis provides comprehensive estimates of the net benefits of Rx fire on subsequent burn severity and smoke PM2.5 emissions in the western US, an empirical basis for evaluating proposed Rx fire expansions, and valuable constraints for future modeling.