Drought is killing an increasing number of trees globally, yet mortality risk remains difﬁcult to predict at ﬁne spatial scales. We sought to identify metrics of living individuals that could be used to estimate mortality risk of Ashe juniper (Juniperus ashei) trees and eventually to estimate the fraction of juniper populations at risk from drought. Ashe juniper is a keystone species in the Edwards Plateau region in central Texas, USA. We analyzed tree rings from both living and dead trees to determine growth rate prior to an historic drought in 2011 and measured morpholog-ical, physiological, and stand-level variables hypoth-esized to link growth rate and mortality risk. Slowly growing trees were disproportionately vulnerable to mortality. Fractional mortality of sampled trees was correlated to the difference between the mean pre-drought basal area increment (BAI) per tree and the predrought BAI of minimally stressed trees growing on deep soil (=BAI90 - BAI). Slowly growing trees had sparsely foliated canopies. BAI90 - BAI was positively correlated to the difference between: (1) leaf area per unit of projected canopy area per tree (LA) and the LA of minimally stressed trees and (2) projected canopy area (CA) and the CA of comparably sized trees. By contrast, there was no correlation between growth of living trees and light interception by neighboring trees, soil depth, or two functional metrics, the stem–leaf W gradient and leaf light use efﬁciency. Mortality risk in Ashe juniper populations can be estimated from nondestructive measurements of leaf and canopy area of individual trees using relationships among risk, growth, and leaf and canopy area.