Abstract Background Snags (standing dead trees) are important biological legacies in forest systems, providing numerous resources as well as a record of recent tree mortality. From 1997 to 2017, we monitored snag populations in drought-influenced mixed-conifer and ponderosa pine (Pinus ponderosa) forests in northern Arizona. Results Snag density increased significantly in both forest types. This increase was driven largely by a pulse in snag recruitment that occurred between 2002 and 2007, following an extreme drought year in 2002, with snag recruitment returning to pre-pulse levels in subsequent time periods. Some later years during the study also were warmer and/or drier than average, but these years were not as extreme as 2002 and did not trigger the same level of snag recruitment. Snag recruitment was not equal across tree species and size classes, resulting in significant changes in species composition and size-class distributions of snag populations in both forest types. Because trees were far more abundant than snags in these forests, the effect of this mortality pulse on tree populations was far smaller than its effect on snag populations. Snag loss rates increased over time during the study, even though many snags were newly recruited. This may reflect the increasing prevalence of white fir snags and/or snags in the smaller size classes, which generally decay faster than snags of other species or larger snags. Thus, although total numbers of snags increased, many of the newly recruited snags may not persist long enough to be valuable as nesting substrates for native wildlife. Conclusions Increases in snag abundance appeared to be due to a short-term tree mortality “event” rather than a longer-term pattern of elevated tree mortality. This mortality event followed a dry and extremely warm year (2002) embedded within a longer-term megadrought. Climate models suggest that years like 2002 may occur with increasing frequency in the southwestern U.S. Such years may result in additional mortality pulses, which in turn may strongly affect trajectories in abundance, structure, and composition of snag populations. Relative effects on tree populations likely will be smaller, but, over time, also could be significant.