Drylands, landscapes characterized by water scarcity, provide ecosystem services that support billions of people worldwide, but these ecosystems have experienced significant degradation in recent decades. In response to degradation processes, drylands often experience long recovery times. The factors influencing the degradation-recovery process of dryland systems also vary across time and space. This variability complicates land management decisions because specific impacts of dryland restoration efforts are not easily identifiable when continuing influence from historical events—legacy effects—could be impacting current vegetation conditions. Insights into the nature of legacy effects in dryland systems could improve ecosystem restoration, but this needs to be done at large spatial scales to capture the complex vegetation processes in drylands. This work provides a new approach to characterize historical vegetation trends in order to better understand dryland landscape processes. We focus on a section of the Chihuahuan Desert in New Mexico, U.S., extracting nearly 40 years of information on vegetation condition from satellite data. We generate classes of vegetation change dynamics over time, demonstrating how this new classification provides novel information on the nature of land cover change in drylands. The knowledge derived from this work on long-term vegetation information can inform restoration-focused land management decisions.