We assess the importance of factors that determine the intensity of UV radiation at the Earth's surface. Among these, atmospheric ozone, which absorbs UV radiation, is of considerable importance, but other constituents of the atmosphere, as well as certain consequences of climate change, can also be major influences. Further, we assess the variations of UV radiation observed in the past and present, and provide projections for the future. Of particular interest are methods to measure or estimate UV radiation at the Earth's surface. These are needed for scientific understanding and, when they are sufficiently sensitive, they can serve as monitors of the effectiveness of the Montreal Protocol and its amendments. Also assessed are several aspects of UV radiation related to biological effects and health. The implications for ozone and UV radiation from two types of geoengineering methods that have been proposed to combat climate change are also discussed. In addition to ozone effects, the UV changes in the last two decades, derived from measurements, have been influenced by changes in aerosols, clouds, surface reflectivity, and, possibly, by solar activity. The positive trends of UV radiation observed after the mid-1990s over northern mid-latitudes are mainly due to decreases in clouds and aerosols. Despite some indications from measurements at a few stations, no statistically significant decreases in UV-B radiation attributable to the beginning of the ozone recovery have yet been detected. Projections for erythemal irradiance (UVery) suggest the following changes by the end of the 21(st) century (2090-2100) relative to the present time (2010-2020): (1) Ozone recovery (due to decreasing ozone-depleting substances and increasing greenhouse gases) would cause decreases in UVery, which will be highest (up to 40%) over Antarctica. Decreases would be small (less than 10%) outside the southern Polar Regions. A possible decline of solar activity during the 21(st) century might affect UV-B radiation at the surface indirectly through changes induced in stratospheric ozone. (2) The projected changes in cloud cover would lead to relatively small effects (less than 3%), except at northern high latitudes where increases in cloud cover could lead to decreases in UVery by up to 7%. (3) Reductions in reflectivity due to the melting of sea-ice in the Arctic would lead to decreases of UVery by up to 10%, while at the margins of the Antarctic the decreases would be smaller (2-3%). The melting of the sea-ice would expose the ocean surface formerly covered by ice to UV-B radiation up to 10 times stronger than before. (4) The expected improvement of air-quality and reductions of aerosols over the most populated areas of the northern hemisphere may result in 10-20% increases in UVery, except over China where even larger increases are projected. The projected aerosol effect for the southern hemisphere is generally very small. Aerosols are possibly the most important factor for future UV levels over heavily populated areas, but their project...