Introduction Background radiation levels in the atmosphere vary in intensity with latitude, altitude and phase of the solar cycle. These background levels are generated primarily by galactic cosmic rays (GCR), consisting of energetic nuclei of all naturally occurring elements, interacting with atmospheric constituents, primarily through atomic and nuclear collisions. Cosmic rays were discovered in 1912, about the same time that aviation began to develop. Over the years, an awareness of the potential hazards faced by flight crews and passengers has developed as air travel becomes more frequent and aircraft fly at higher altitudes for longer periods of time. Estimates of potential exposures of aircrew and frequent travellers, validated by on-board measurements carried out by carriers in the United States, Canada and Europe, indicate that these groups sometimes receive exposures above the limits for the general public, and should be categorized as occupationally exposed.Recent European directives require that aircrew liable to receive 1 mSv y-1 or more from their flight duties must have their exposures assessed . For civilian aircrew this can be achieved by matching computer estimates of route doses with flight crew rosters to assess exposures. For military crews, however, their irregular flight patterns make computer estimates of doses impractical. Stokes and Talbot, in the accompanying article in this issue , explore this problem and suggest that actual dosimeter measurements may be more easily obtained and be more reliable. In the article, these and related issues are briefly discussed and placed in the broader context of radiation protection issues for atmospheric flight. Those readers interested in more in-depth treatments of these issues are referred to several recently published reports and conference proceedings [3-5]. Reference , which contains the Proceedings of the 1998 Annual Meeting of the National Council on Radiation Protection and Measurements, is particularly helpful because it is recent and has great depth and breadth of coverage of the topic of aircrew exposure and risk. This editorial begins with an introduction to the radiation environment in the atmosphere, its sources, and associated exposure rates. This is followed by a brief discussion of the implications of these exposures on present and future air travel. Atmospheric radiation environment The background radiation levels in the atmosphere are primarily generated by galactic cosmic rays, which consist of energetic nuclei of all naturally occurring elements interacting with atmospheric constituents. Nearly 90 per cent of the incident particles are energetic protons, about 10 per cent are alpha particles, and the remaining 1-2 per cent are composed of nuclei as heavy as uranium. These incident GCR particles have broad energy spectra, with peak fluxes at kinetic energies of the order of several GeV/nucleon, and maximum energies exceeding 1018 eV. The GCR intensity in the inner solar system varies with the phase of the approximate 11-year solar cycle. During periods when the solar plasma emitted into space is at its maximum, the GCR intensity is reduced as the incoming particles, especially those with energies below several hundred MeV/nucleon, are deflected by the magnetic fields frozen in the plasma streaming away from the sun. Conversely, when solar activity is at its quietest (solar minimum), GCR intensity is a maximum since the incoming ions are deflected less by the reduced emission of solar plasma. In the vicinity of the Earth the incoming GCR ions are further affected by the Earth's magnetic field. Near the geomagnetic equator, most incoming ions are deflected back into space before entering Earth's atmosphere. Only very energetic ions are able to reach the atmosphere and interact. As one moves from the geomagnetic equator towards the poles, the Earth's magnetic field lines become more parallel to the trajectories of the incoming ions and the fluxes that penetrate into the atmosphere increase. Hence, the intensity of radiation in the atmosphere increases with increasing latitude. Near the poles ion trajectories are nearly parallel to the magnetic field lines and are largely unaffected by them. Therefore, the incident GCR fluxes approach those found in deep space. Another source of ions, which can be very intense and sporadic in occurrence, is a large solar energetic particle event. This occurs when localized, short-term activity on the sun results in large fluxes of high-energy particles, mainly protons. Sometimes these events are accompanied by geomagnetic storms, which reduce the intensity of Earth's magnetic field, thereby increasing the fluxes of particles entering the atmosphere, particularly at high latitudes. When these energetic ions enter Earth's atmosphere, they undergo nuclear collisions with atmosphere constituents. The nuclear spallation or fragmentation (breakup) processes that occur in these collisions result in the production of energetic secondary radiations, especially neutrons. As these secondaries penetrate deeper into the atmosphere, they are attenuated. Therefore, radiation exposure rates in the atmosphere vary with phase of the solar cycle (greatest during solar minimum when GCR fluxes are highest), altitude (generally increase with altitude) and latitude (higher at higher latitudes). At sea level the exposure rate is ~0.03 μSv h-1. At typical commercial jet aircraft altitudes of 9-12 km (30 000-40 000 ft) the exposure rates increase to ~5-10 μSv h-1 . At Concorde flight altitudes, and those of various proposed high speed commercial passenger aircraft (about 18-20 km or 59 000-65 000 ft), the estimated dose equivalent rates are ~10-20 μSv h-1 , in good agreement with the measurements on Concorde reported by Davies  in 1993. Recent measurements on Concorde, for the month of October 1996, are 14.1 μSv h-1 . At flight altitudes above 3 km (10 000 ft) the dominant contributor to the effective dose is from secondary neutrons . Implications for current and future air travel In a 1993 UNSCEAR report , the annual average effective dose to aircrew was reported as 3 mSv y-1. This is slightly higher than the value of 2.9 mSv y-1 reported for nuclear workers. Within the United States of America, Federal Aviation Administration estimates of effective doses for crews on various routes within the USA and between the USA and Europe or Asia range from 0.03 to 0.63 mSv per 100 flight hours . Hence, aircrew flying 700 hours annually could expect an effective dose of 0.2-4.4 mSv. Over a 30 y career, the cumulative effective doses would range from 0.06-0.13 Sv. The estimated increase in lifetime risk of fatal cancer because of these exposures is less than 0.5 per cent . For Canadian aircrew, a study led by the Royal Military College of Canada  indicates that most Canadian-based domestic and international aircrew will exceed the annual limit from ICRP Publication 60  of 1 mSv y-1 for the public, but should not exceed the occupational limit of 20 mSv y-1. The maximum effective dose obtained in the study, for a flight attendant with over 1100 flight hours, was 4.9 mSv. In Europe, measurement programmes have been carried out with several international airlines, including Aer Lingus, Alitalia, British Airways, Lufthansa and Scandinavian Airlines, over a large number of commercial routes . Ambient dose equivalent rates ranging from 2-10 μSv h-1 were measured. Based upon these results, annual effective doses may exceed the annual limit of 1 mSv for the public for many commercial flight crews, especially those on long haul, international routes. Hence, these flight crews should be treated as radiation workers and their exposures assessed. For a 30 y career, effective doses of ~0.1 Sv are expected. The increased risk of developing a fatal cancer for this level of exposure is ~0.5 per cent, which should be contrasted with the ~25 per cent risk of cancer mortality within the USA from all sources. A special category of exposed workers is the pregnant crewmember. Exposure limits for these individuals are less than 1 mSv . Some carriers, including British Airways, ground these crewmembers for the duration of their pregnancies. Another exposed population, which has received increased attention in recent years, is the frequent business traveller. As discussed by Bagshaw , exposures to these individuals exceeding 1 mSv y-1 are possible. Consequently, classifying those business travellers who are frequent fliers as radiation workers has been suggested. Again, however, the risk of developing a fatal cancer, as a result of these levels of exposure, is very low (less than 0.5 per cent). In the future, supersonic aircraft will be developed that fly higher, faster, and farther than existing aircraft. Thus, future increases in effective doses from radiation in the atmosphere are expected for aircrew and frequent fliers. Additional measures to monitor and limit exposures may be necessary to comply with requirements to keep exposures as low as reasonably achievable (ALARA). Based upon measurements made at high altitudes (~20 km) during the joint NASA/Department of Energy Atmospheric Ionizing Radiation (AIR) project by instruments flown on a NASA ER-2 aircraft, annual effective doses to aircrew and frequent fliers could approach the 20 mSv y-1 limit for radiation workers. Accrual of such annual exposures over a 30 y career could increase the risk of developing a fatal cancer by several per cent in the exposed populations, unless actions to reduce the accrued effective doses are implemented in accordance with ALARA requirements. Finally, we note that military aircrew are also expected to receive annual effective doses that exceed 1 mSv. Hence, these aircrew are also radiation workers. Assessing doses to military crews using a combination of crew rosters and route doses, as is often done for commercial aircrew, is impractical since military flight patterns are highly variable. Therefore, monitoring of military crews using dosimeters, as proposed by Stokes and Talbot , appears to be the best option for assessing radiation risks and demonstrating compliance with applicable exposure limits.