Over the past century human life expectancy has risen substantially around the world, and longevity in the modern world is largely determined by the age at which adults die. There is an ongoing, longstanding trend to higher average ages at death for both sexes in most countries. Accompanying this trend has been a decrease in the dispersion of the age of adult death which has slowed in most countries in the past five decades. Mortality forecasts can be made using the Lee-Carter method and provide stochastic projections of average longevity. We may think of the dispersion in the forecasted averages as a form of systematic risk that is important in pricing longevity risk. The dispersion around the average age at death is a form of individual risk. This individual risk certainly does matter to individuals making decisions about consumption, savings, bequests and so on. In particular it matters to economic aggregates such as population wealth, health, and savings, and to the fiscal flows in private and public pensions. Realistic models of annuity valuation are likely to be nonlinear and so annuity values are also likely to depend on the individual component of longevity risk. Analysis of the distribution of individual risk in subgroups (e.g., with high or low educational attainment) suggests that selection for low risk subgroups may be difficult to achieve.