Summary Microrefugia are important for supporting populations during periods of unfavourable climate change and in facilitating rapid migration as conditions ameliorate. With ongoing anthropogenic climate change, microrefugia could have an important conservation value; however, a simple tool has not been developed and tested to predict which settings are microrefugial. We provide a tool based on terrain ruggedness modelling of individual catchments to predict Andean microrefugia. We tested the predictions using nine Holocene Polylepis pollen records. We used the mid‐Holocene dry event, a period of peak aridity for the last 100 000 yr, as an analogue climate scenario for the near future. The results suggest that sites with high terrain rugosity have the greatest chance of sustaining mesic conditions under drier‐than‐modern climates. Fire is a feature of all catchments; however, an increase in fire is only recorded in settings with low rugosity. Owing to rising temperatures and greater precipitation variability, Andean ecosystems are threatened by increasing moisture stress. Our results suggest that high terrain rugosity helps to create more resilient catchments by trapping moisture through orographic rainfall and providing firebreaks that shelter forest from fire. On this basis, conservation policy should target protection and management of catchments with high terrain rugosity.