Trait-based approaches are increasingly recognized as a tool for understanding ecosystem re-assembly and function under intensifying global change. Here we synthesize trait-based research globally (N=865 studies) to examine the contexts in which traits may be used for global change prediction. We find that exponential growth in the field over the last decade remains dominatedby descriptive studies of terrestrial plant morphology, highlighting significant opportunities to expand traits-based thinking across systems and taxa. Very few studies (<3%) focus onpredicting ecological effects of global change, mostly in the past five years and via singular traitsthat mediate abiotic limits on species distribution. Beyond organism size (the most examinedtrait), we identify >2,500 other morphological, physiological, behavioural, and life history traitsknown to mediate environmental filters of species’ range and abundance as candidates for futurepredictive global change work. Though uncommon, spatially-explicit process models—whichmechanistically link traits to changes in organism distributions and abundance—are among themost promising frameworks for holistic global change prediction at scales relevant forconservation decision-making. Further progress towards traits-based forecasting requiresaddressing persistent barriers including (1) matching scales of multivariate trait and environmentdata to focal processes disrupted by global change, and (2) propagating variation in trait andenvironmental parameters throughout process model functions using simulation.