Accuracy in seasonal-to-interannual climate forecasts for the United States (US) remains a challenge. This despite advances in understanding sources of climate variability and predictability as well as improvements in prediction tools. Our use of the tools has greatly improved in the past decade with the implementation of robust model bias correction and multi-modeling strategies. Furthermore, validation measures have become more sophisticated, rating the performance of forecast systems in a manner more consistent with the probabilistic world they describe. Still, further room for improvement exists. This article outlines the current practices of seasonal-to-interannual climate prediction: current understanding of the sources of variability, the tools used to predict it, common methodologies applied to those tools to produce forecasts, and relevant verification analyses with which to judge the performance of the forecasts. These are forecasts of opportunity, which if used prudently have potential to benefit decision-making. Background Before discussing current prediction practices and their accuracies, it is important to distinguish between prediction and predictability itself. The latter is a physical characteristic of the natural system, and is not altered by forecasting methodologies. The tools used to make forecasts are often employed in determining the theoretical limit of predictability, judging the model against itself, and as such predictability estimates can indeed change (for non-physical reasons) as models evolve. Nonetheless, it is often of interest to know how the current skill levels differ from the existing theoretical limits because such knowledge guides expectations for the skill impacts of improved practices. However, given the indeterminate nature of predictability estimates, this report focuses on skill estimates obtained by comparing model-derived forecasts with the observed climate, emphasizing seasonal mean surface temperature and precipitation variations over the US. Attributable causes of US seasonal climate variability Understanding US seasonal climate variability is essential for exposing the sources of its predictability. Seasonal forecasting (when done at the minimal 15-day lead times beyond which deterministic atmospheric predictions are skillful) is effectively the practice of predicting the climate signal due to external forcings. These forcings include anomalous sea surface temperature (SST), soil moisture, sea ice, and chemical constituents. The resulting climate predictability is known as predictability of the "second kind" or 2-tier, arising from the influence of specified boundary conditions on the atmosphere. For seasonal prediction practices using fully coupled Earth System models, the notion of such a 2-tiered system with external forcings vanishes, and predictability is of the "first kind" (ie 1-tier) arising solely from the initial Earth System conditions. It is important to note here that for seasonal prediction, longer-term changes of external forcing that is affecting the climate system, especially increasing greenhouse gasses, may be considered constant