We review the ideas behind the pattern scaling technique, and focus on its value and limitations given its use for impact assessment and within integrated assessment models. We present estimates of patterns for temperature and precipitation change from the latest transient simulations available from the Coupled Model Inter-comparison Project Phase 5 (CMIP5), focusing on multi-model mean patterns, and characterizing the sources of variability of these patterns across models and scenarios. The patterns are compared to those obtained from the previous set of experiments, under CMIP3. We estimate the significance of the emerging differences between CMIP3 and CMIP5 results through a bootstrap exercise, while also taking into account the fundamental differences in scenario and model ensemble composition. All in all, the robustness of the geographical features in patterns of temperature and precipitation, when computed as multi-model means, is confirmed by this comparison. The intensity of the change (in both the warmer and cooler areas with respect to global temperature change, and the drier and wetter regions) is overall heightened per degree of global warming in the ensemble mean of the new simulations. The presence of stabilized scenarios in the new set of simulations allows investigation of the performance of the technique once the system has gotten close to equilibrium. Overall, the well established validity of the technique in approximating the forced signal of change under increasing concentrations of greenhouse gases is confirmed.