Societal adaptation strategies to climate change call for climate projections over the next few decades, particularly at the regional scale. This raises the question to what extent climate evolution is predictable. Past climate variations offer a wealth of case studies to test climate predictability on a variety of timescales. This could be based on the statistical analysis of climate response to a known forcing, for example solar and volcanic forcings, or perturbations of the carbon cycle or the large-scale ocean circulation. Another dynamical system approach may involve the analysis of critical slowdown before state transitions. Paleoclimate information is moreover expected to allow investigations of the dependency of climate predictability on the background climate state. We invite contributions from all areas of paleoclimate research, covering statistical analyses of paleoclimatic records, as well as theoretical and modeling approaches.