Uncertainties about future climate conditions underscores the need to better understand genotype x environment (GxE) interactions to support resilient coffee cultivar adoption and breeding. We Assessed yield and coffee leaf rust severity using a unique worldwide trial of 31 Coffea arabica L. varieties evaluated across 29 sites in 18 countries, and linked varietal performance to global climate gradients. A two-stage analysis using linear mixed-effects models with a two-factor analytic (FA) structure was applied to estimate GxE variance-covariance for both traits. Partial Least Squares Regression linked the latent FA axes to climate gradients. The FA(2) model explained 58 percent of the GxE variance for yield and 67 percent for rust severity. For yield, fa1 representing performance was not associated with climate. fa2, representing dynamic stability, was correlated (p < 0.10) with a gradient from cool environments with high diurnal temperature range to warmer environments with lower temperature range. For rust severity, both fa axes were climate-associated (p < 0.10), reflecting gradients related to temperature, rainfall patterns, humidity, and diurnal versus seasonal temperature variation. Across environments, EC16 and H1 Centroamericano (F1 hybrids), followed by Parainema (sarchimor), showed the best combined yield and rust performance, but also exhibited specific adaptation with limited yield stability. A group of introgressed varieties demonstrated a favorable performance-stability tradeoff, while pure arabica lines showed the lowest performance and stability. These exploratory results highlight climate-linked drivers of GxE interaction and provide insights to guide variety recommendations and future coffee breeding strategies.