Fighters were typically armed with guns only for air to air combat up through the late 1950s, though unguided rockets for mostly air to ground use and limited air to air use were deployed in WWII. From the late 1950s forward guided missiles came into use for air to air combat. Throughout this history fighters which by surprise or maneuver attain a good firing position have achieved the kill about one third to one half the time, no matter what weapons were carried. The only major historic exception to this has been the low effectiveness shown by guided missiles in the first one to two decades of their existence. From WWI to the present, fighter aircraft have featured machine guns and automatic cannons as weapons, and they are still considered as essential back-up weapons today. The power of air-to-air guns has increased greatly over time, and has kept them relevant in the guided missile era. In WWI two rifle (approximately 0.30) caliber machine guns was the typical armament, producing a weight of fire of about 0.4 kg (0.88 lb) per second. In WWII rifle caliber machine guns also remained common, though usually in larger numbers or supplemented with much heavier 0.50 caliber machine guns or cannons. The standard WWII American fighter armament of six 0.50-cal (12.7mm) machine guns fired a bullet weight of approximately 3.7 kg/sec (8.1 lbs/sec), at a muzzle velocity of 856 m/s (2,810 ft/s). British and German aircraft tended to use a mix of machine guns and autocannon, the latter firing explosive projectiles. Later British fighters were exclusively cannon-armed, the US were not able to produce a reliable cannon in high numbers and most fighters remained equipped only with heavy machine guns despite the US Navy pressing for a change to 20mm.
Post war 20-30 mm revolver cannon and rotary cannon were introduced. The modern M61 Vulcan 20 mm rotary cannon that is standard on current American fighters fires a projectile weight of about 10 kg/s (22 lb/s), nearly three times that of six 0.50-cal machine guns, with higher velocity of 1,052 m/s (3450 ft/s) supporting a flatter trajectory, and with exploding projectiles. Modern fighter gun systems also feature ranging radar and lead computing electronic gun sights to ease the problem of aim point to compensate for projectile drop and time of flight (target lead) in the complex three dimensional maneuvering of air-to-air combat. However, getting in position to use the guns is still a challenge. The range of guns is longer than in the past but still quite limited compared to missiles, with modern gun systems having a maximum effective range of approximately 1,000 meters. High probability of kill also requires firing to usually occur from the rear hemisphere of the target. Despite these limits, when pilots are well trained in air-to-air gunnery and these conditions are satisfied, gun systems are tactically effective and highly cost efficient. The cost of a gun firing pass is far less than firing a missile, and the projectiles are not subject to the thermal and electronic countermeasures than can sometimes defeat missiles. When the enemy can be approached to within gun range, the lethality of guns is approximately a 25% to 50% chance of "kill per firing pass".
The range limitations of guns, and the desire to overcome large variations in fighter pilot skill and thus achieve higher force effectiveness, led to the development of the guided air-to-air missile. There are two main variations, heat-seeking (infrared homing), and radar guided. Radar missiles are typically several times heavier and more expensive than heat-seekers, but with longer range, greater destructive power, and ability to track through clouds.
The highly successful AIM-9 Sidewinder heat-seeking (infrared homing) short-range missile was developed by the United States Navy in the 1950s. These small missiles are easily carried by lighter fighters, and provide effective ranges of approximately 10 to 35 km (~6 to 22 miles). Beginning with the AIM-9L in 1977, subsequent versions of Sidewinder have added all-aspect capability, the ability to use the lower heat of air to skin friction on the target aircraft to track from the front and sides. The latest (2003 service entry) AIM-9X also features "off-boresight" and "lock on after launch" capabilities, which allow the pilot to make a quick launch of a missile to track a target anywhere within the pilot's vision. The AIM-9X development cost was U.S. $3 billion in mid to late 1990s dollars, and 2015 per unit procurement cost is $0.6 million each. The missile weighs 85.3 kg (188 lbs), and has a maximum range of 35 km (22 miles) at higher altitudes. Like most air-to-air missiles, lower altitude range can be as limited as only about one third of maximum due to higher drag and less ability to coast downward.
The effectiveness of infrared homing missiles was only 7% early in the Vietnam War, but improved to approximately 15%–40% over the course of the war. The AIM-4 Falcon used by the USAF had kill rates of approximately 7% and was considered a failure. The AIM-9B Sidewinder introduced later achieved 15% kill rates, and the further improved AIM-9D and J models reached 19%. The AIM-9G used in the last year of the Vietnam air war achieved 40%. Israel used almost totally guns in the 1967 Six-Day War, achieving 60 kills and 10 losses. However, Israel made much more use of steadily improving heat-seeking missiles in the 1973 Yom Kippur War. In this extensive conflict Israel scored 171 of 261 total kills with heat-seeking missiles (65.5%), 5 kills with radar guided missiles (1.9%), and 85 kills with guns (32.6%). The AIM-9L Sidewinder scored 19 kills out of 26 fired missiles (73%) in the 1982 Falklands War. But, in a conflict against opponents using thermal countermeasures, the United States only scored 11 kills out of 48 fired (Pk = 23%) with the follow-on AIM-9M in the 1991 Gulf War.
Radar guided missiles fall into two main missile guidance types. In the historically more common semi-active radar homing case the missile homes in on radar signals transmitted from launching aircraft and reflected from the target. This has the disadvantage that the firing aircraft must maintain radar lock on the target and is thus less free to maneuver and more vulnerable to attack. A widely deployed missile of this type was the AIM-7 Sparrow, which entered service in 1954 and was produced in improving versions until 1997. In more advanced active radar homing the missile is guided to the vicinity of the target by internal data on its projected position, and then "goes active" with an internally carried small radar system to conduct terminal guidance to the target. This eliminates the requirement for the firing aircraft to maintain radar lock, and thus greatly reduces risk. A prominent example is the AIM-120 AMRAAM, which was first fielded in 1991 as the AIM-7 replacement, and which has no firm retirement date as of 2016. The current AIM-120D version has a maximum high altitude range of greater than 160 km (>99 miles), and cost approximately $2.4 million each (2016). As is typical with most other missiles, range at lower altitude may be as little as one third that of high altitude.
In the Vietnam air war radar missile kill reliability was approximately 10% at shorter ranges, and even worse at longer ranges due to reduced radar return and greater time for the target aircraft to detect the incoming missile and take evasive action. At one point in the Vietnam war, the U.S. Navy fired 50 AIM-7 Sparrow radar guided missiles in a row without a hit. Between 1958 and 1982 in five wars there were 2,014 combined heat-seeking and radar guided missile firings by fighter pilots engaged in air-to-air combat, achieving 528 kills, of which 76 were radar missile kills, for a combined effectiveness of 26%. However, only four of the 76 radar missile kills were in the beyond-visual-range mode intended to be the strength of radar guided missiles. The United States invested over $10 billion in air-to-air radar missile technology from the 1950s to the early 1970s. Amortized over actual kills achieved by the U.S. and its allies, each radar guided missile kill thus cost over $130 million. The defeated enemy aircraft were for the most part older MiG-17s, −19s, and −21s, with new cost of $0.3 million to $3 million each. Thus, the radar missile investment over that period far exceeded the value of enemy aircraft destroyed, and furthermore had very little of the intended BVR effectiveness.
However, continuing heavy development investment and rapidly advancing electronic technology led to significant improvement in radar missile reliabilities from the late 1970s onward. Radar guided missiles achieved 75% Pk (9 kills out of 12 shots) in operations in the Gulf War in 1991. The percentage of kills achieved by radar guided missiles also surpassed 50% of total kills for the first time by 1991. Since 1991, 20 of 61 kills worldwide have been beyond-visual-range using radar missiles. Discounting an accidental friendly fire kill, in operational use the AIM-120D (the current main American radar guided missile) has achieved 9 kills out of 16 shots for a 56% Pk. Six of these kills were BVR, out of 13 shots, for a 46% BVR Pk. Though all these kills were against less capable opponents who were not equipped with operating radar, electronic countermeasures, or a comparable weapon themselves, the BVR Pk was a significant improvement from earlier eras. However, a current concern is electronic countermeasures to radar missiles, which are thought to be reducing the effectiveness of the AIM-120D. Some experts believe that as of 2016 the European Meteor missile, the Russian R-37M, and the Chinese PL-15 are more resistant to countermeasures and more effective than the AIM-120D.
Now that higher reliabilities have been achieved, both types of missiles allow the fighter pilot to often avoid the risk of the short-range dogfight, where only the more experienced and skilled fighter pilots tend to prevail, and where even the finest fighter pilot can simply get unlucky. Taking maximum advantage of complicated missile parameters in both attack and defense against competent opponents does take considerable experience and skill, but against surprised opponents lacking comparable capability and countermeasures, air-to-air missile warfare is relatively simple. By partially automating air-to-air combat and reducing reliance on gun kills mostly achieved by only a small expert fraction of fighter pilots, air-to-air missiles now serve as highly effective force multipliers.