Any risk-based assessment of future sea level must include at least the following two statistical descriptions of the future: (1) the uncertain projection of sea-level rise and (2) the distribution of extreme sea levels due to processes such as tides and storm surges. The combination of these statistics indicates the expected change in the frequency of ocean flooding events at a given height. A simple implementation, which involves a straightforward convolution of the statistical distributions, involves the assumption that the extremes (2) will not change relative to mean sea level, and any given sea-level rise simply raises the entire extremes distribution without changing its shape. If, in addition, it is assumed (often quite appropriately) that the extremes distribution is Gumbel, then the solution shows that, for a given (uncertain) sea-level rise, the frequency of flooding events at any height is increased by a constant factor. A consequence of this is the definition of a single "allowance" for any given sea-level rise; any item of infrastructure raised by this allowance would experience the same frequency of flooding events under sea-level rise as it would without the allowance and without sea-level rise. One feature of the allowance is that it depends on the shape of the uncertainty distribution of the sea-level projection; a non-normal distribution with a fat high tail can significantly increase the allowance. I will discuss current estimates of this distribution and its high tail, how we might expect this distribution to change with new knowledge, and implications for planning for future sea-level rise.