The U.S. Air Force started its first military space program in 1956. Known as Weapons System 117L (or more simply, WS-117L), it was an effort to develop a television-based reconnaissance satellite for spying on the Soviet Union. The program limped along under funded until after the Soviet launch of Sputnik, in October 1957, when it shifted into high gear. At this point, the U.S. Air Force began developing two types of satellites. One was a system, soon named Samos, that would take a picture on film and develop it in orbit, scanning it like a computer scanner and relaying the data to the ground over a radio link. This system was slow, and the satellites could transmit only a few dozen images a day, which was too little to have any value. The Samos readout system was cancelled in the early 1960s.

 
First imagery taken by CORONA - Mys Shmidta Air Field, USSR 18 Aug 1960.

The other system, which was led by the Central Intelligence Agency (CIA) with substantial Air Force participation, was named CORONA in February 1958 (the names of U.S. intelligence satellites are always written in all upper-case letters). This system used a panoramic camera on a stabilized satellite to take a long, thin image on a 70-millimeter strip of film.

CORONA recovery sequence.

The film wound on a reel inside a small bucket-shaped re-entry vehicle and re-entered the atmosphere, where it was snatched out of the air by a passing airplane while hanging from its parachute. CORONA could see objects as small as six to nine feet (two to three meters) on a side. Only the film was saved; each satellite was used for only a single mission. The first successful mission was in August 1960 and more than 120 satellites were launched until the early 1970s, when the system was replaced by a much more massive satellite commonly referred to as the “Big Bird,” but actually named the KH-9 HEXAGON.

 
CORONA image showing transport vehicles and bombers.

Beginning in 1960, the Air Force began developing a new, high-resolution satellite system called the KH-7 GAMBIT. This satellite could take very powerful images of a small area on the ground, allowing an intelligence analyst to look at targets as small as 1.5 feet (0.5 meters) across. By 1967, an even more powerful version entered service, allowing experts to see objects as small as about three inches. This was the most powerful satellite ever operated by any nation, but it did not allow operators to read license plates, like Hollywood satellites could.

 
The CORONA cameras.

The Soviet Union also started its own reconnaissance satellite program in the late 1950s. This effort, known as Zenit, used a modified version of the Vostok piloted spacecraft, equipped with several cameras that took still images. The entire large re-entry vehicle would return to Earth with both film and cameras, allowing the cameras to be reused. The Soviets also developed a higher-resolution system called Yantar, versions of which are still in use today.

After Sputnik, the U.S. Navy began developing a small satellite for intercepting Soviet radar transmissions. Several of these satellites, named GRAB, and later DYNO, were launched starting in 1960 and followed by a better series of satellites named POPPY. They intercepted the electronic signals emitted by ground-based aircraft search radars, allowing the Navy and the Air Force Strategic Air Command to locate the radar positions and develop ways of avoiding or jamming them. A more capable system, called PARCAE, used groups of small satellites. It could locate ships at sea based upon their radar and radio transmissions by triangulating their position.

Starting in 1961, all of these intelligence collection satellites were operated by a highly secret agency within the Department of Defence called the National Reconnaissance Office, or NRO. The NRO was not even publicly acknowledged until 1992.

Throughout the 1960s, the United States and the Soviet Union operated much bigger and better versions of their film-return satellites and signals intelligence satellites. In the late 1960s the U.S. Air Force began operating a communications intelligence satellite named CANYON that intercepted Soviet telephone transmissions. Follow-on versions of this satellite, named VORTEX and MERCURY, continued to operate into the 21st century. In the early 1970s, the CIA deployed a satellite named AQUACADE for intercepting the faint radio signals from Soviet missile tests, using a giant mesh dish for detecting the electronic whispers. Modern versions named MERCURY and ORION were launched in the 1980s, and even more upgraded versions were launched in the 1990s. These satellites are believed to have truly massive antennas that unfurl in geosynchronous orbit.

The one major limitation of the film-return satellites like CORONA, HEXAGON, GAMBIT and Zenit and Yantar is that it could take days for their images to reach the eyes of an intelligence expert. Both countries investigated whether piloted reconnaissance satellites could partly fulfil this role. The U.S. Air Force's Manned Orbiting Laboratory (MOL) was designed to assess the capabilities and potential of military operations in space but was abandoned in 1969, before it ever flew. The Soviet Union launched several Almaz military space stations during the 1970s before abandoning its effort. Both countries realized that humans and sensitive optics were a bad combination.

Throughout the 1960s both the U.S. Air Force and the CIA sought to develop a “real-time” imaging satellite system that could transmit its images to the ground as fast as it took them. The CIA sponsored the development of what became the Charge Coupled Device, or CCD, like that used in modern digital cameras. In December 1976, the NRO launched the first satellite using this new technology, called the KH-11 KENNAN. The KH-11 allowed images to reach an intelligence expert, or the desk of the President of the United States, within minutes. It too revolutionized intelligence collection and by the 1980s the United States no longer launched film-return satellites. These satellites look much like NASA's Hubble Space Telescope, with a large mirror inside a long tube, except they point at the Earth rather than the stars. They can probably see objects about the size of a softball. Versions of the KH-11 equipped with some infrared capability (later called CRYSTAL and now called by some highly-secret code name) continue to operate today and their images are vital to U.S. military operations.

 
KH-11 photo of the Shifa Pharmaceutical Plant, Sudan.

The Soviet Union continued to lag behind the United States and did not orbit its own real-time reconnaissance satellite until the 1980s. After the Cold War ended, Russian military space launches diminished to a trickle and they launched fewer reconnaissance satellites. Surprisingly, the Russians still launch the occasional film-return satellite, possibly because their real-time satellite, although fast, cannot take high-quality photos.

The final major limitation to imaging satellites was clouds. In the 1960s, the U.S. Air Force launched a radar satellite named QUILL, capable of peering through clouds, but it could not see much detail. In the late 1980s, the NRO launched the first of a new class of large imaging satellites, named LACROSSE (and later ONYX). These satellites probably see images on the ground about three feet (one meter) across.