Human activity has changed our climate over the past century, and further change is inevitable over the next several decades even if we take strong mitigation actions. Thus it is imperative to improve our ability to predict how the climate with evolve over the coming decades in order to inform the adaptation decisions that we must collectively make. In order to advance our capacity to predict climate, and to know whether our predictions are successful, it will be necessary to better understand how human influences have affected the climate on global, regional and local scales over the recent past, to develop means to use this understanding to better constrain predictions of the next few decades, and to grapple with the challenge of assessing the skill of decadal climate predictions. This session will bring together practitioners from the detection and attribution community on the one hand, and the climate prediction community on the other, in order to explore these issues.