Today solar panels are the most common technique to supply power to satellites. Solar panels will work as long as the power demand of the satellite is limited and the satellite can be equipped with enough panels, and kept in an orbit that allows enough sunlight to hit the panels. There are various types of space missions that do not fulfil these criteria. With nuclear power these types of missions can be powered regardless of the sunlight and as early as 1961 the first satellite with a nuclear power source was placed in orbit. Out of seventy known space missions that has made use of nuclear power, ten have had some kind of failure. In no case has the failure been associated with the nuclear technology used. This report discusses to what degree satellites with nuclear power are a source for potential radioactive contamination of Swedish territory. It is not a discussion for or against nuclear power in space. Neither is it an assessment of consequences if radioactive material from a satellite would reach the earth's surface. Historically two different kinds of Nuclear Power Sources (NPS) have been used to generate electric power in space. The first is the reactor where the energy is derived from nuclear fission of 235U and the second is the Radioisotope Thermoelectric Generator (RTG) where electricity is generated from the heat of naturally decaying radionuclides. NPS has historically only been used in space by United States and the Soviet Union (and in one failing operation Russia). Nuclear Power Sources have been used in three types of space objects: satellites, space probes and moon/Mars vehicles. USA has launched one experimental reactor into orbit, all other use of NPS by the USA has been RTG:s. The Soviet Union, in contrast, only launched a few RTG:s but nearly forty reactors. The Soviet use of NPS is less transparent than the use in USA and some data published on Soviet systems are more or less well substantiated assessments. It is likely that also future space probes, moon and Mars vehicles will be using NPS. Besides the more established users of NPS in space, USA and Soviet Union (today Russia), it is possible that we in a not to distant future will see use of NPS in space by China, India and maybe also ESA (European Space Agency). In 1992 the United Nations General Assembly adopted a resolution regarding principles for the use of NPS in space. The resolution consists of eleven points regarding definitions and usage of NPS in space, and how to handle notification and compensation in case of damages due to a failure involving a satellite with an NPS. The probability of radioactive fallout in Sweden following an incident with a NPS-equipped satellite is very small. Due to the fact that everything placed in orbit around Earth sooner or later will re-enter, it is not possible to use probability of re-entry at any time as a measure of risk. Instead the measure Probability of re-entry within 100 year has been chosen. If the routine use of NPS in Low Earth Orbit (LEO) is not taken up again two cases can be defined: - Within about 3 000 years all satellites stored in Nuclear Safe Orbit (NSO) will de-orbit and re-enter the Earth atmosphere. One satellite, Triad OI-IX is in orbit at a lower altitude, and will thus de-orbit earlier. The probability that it does re-enter within 100 years from now is so small that a quantitative measure is deemed not to be meaningful. - There is a risk of a launch failure involving a satellite or space probe with a NPS, with a risk of fallout in Sweden. This is not a large risk, but it is orders of magnitude higher than the probability of a satellite that now is in NSO will end up in Sweden within 100 years. If the routine use of NPS in LEO is re-established, the probabilities above are no longer valid.