Railguns are being pursued as weapons with projectiles that do not contain explosives, but are given extremely high velocities: 3500 m/s or more (for comparison, the M16 rifle has a muzzle speed of 975-1025 m/s), which would make their kinetic energy equal or superior to the energy yield of an explosive-filled shell of greater mass. This would allow more ammunition to be carried and eliminate the hazards of carrying explosives in a tank or naval weapons platform. Also, by firing at higher velocities railguns have greater range, less bullet drop and less wind drift, bypassing the inherent cost and physical limitations of conventional guns - "the limits of gas expansion prohibit launching an unassisted projectile to velocities greater than about 1.5 km/s and ranges of more than 50 miles [80 km] from a practical conventional gun system.^"

Although full scale guns have been built and fired, including a very successful 90 mm bore, 9 MJ kinetic energy gun developed by DARPA, they all suffer from extreme rail damage and need to be serviced after every shot. Rail and insulator ablation issues still need to be addressed before railguns can start to replace conventional weapons. Probably the most successful system was built by the UK's Defence Research Agency at Dundrenna Range in Kirkcudbright, Scotland. This system has now been operational for over 10 years as an associated flight range for internal, intermediate, external and terminal ballistics, and is the holder of several mass and velocity records.

The United States military is funding railgun experiments. At the University of Texas at Austin Institute for Advanced Technology, military railguns capable of delivering tungsten armor piercing bullets with kinetic energies of nine million joules have been developed. Nine million joules is enough energy to deliver 2 kg of projectile at 3 km/s - at that velocity a tungsten or other dense metal rod could penetrate a tank.

Due to the very high muzzle velocity that can be attained with railguns, there is interest in using them to shoot down high-speed missiles.

Naval forces are also interested in railgun research. Current ship guns store their explosive shells in a large magazine underneath the gun. If a shell from a hostile happens to penetrate into the armoury and explode, it is quite likely to cause all of the shells in the magazine to detonate, usually destroying the ship (this is accepted as causing the sinking of the Hood when she fought Bismarck during WWII, and also the sinking of USS Arizona (BB-39) during the attack upon Pearl Harbor). However if the ship is instead equipped with railguns, the magazine would need only to store the non-explosive tungsten bullets. Additionally, the compact railgun projectiles would require less space to store than the shells used for current guns. Electricity for the railgun could be supplied from an on-board compulsator, which in turn could be powered by the ship's engines. However the main advantage for naval forces is range; the US Navy plans to deploy railguns with ranges over 250 miles (400 km) on naval vessels as early as 2011.

Research into tank-portable railguns is in very early stages; high-velocity kinetic energy projectiles are becoming more important as the primary means of penetrating reactive armor.

Man-portable railguns will not be revolutionary weapons; if power supply technology ever allows a railgun small enough to be carried then rail-handguns will probably only be able to fire projectiles at speeds similar to those currently achieved with chemical propellants. The simple reason is that the destructive power of a handgun or long gun is limited as much by recoil as anything else; it is quite possible to build a handgun that fires 20 mm cannon shells, but the recoil would make it impossible to aim or fire safely.

However, the ability for the soldier to adjust the muzzle velocity gives the potential for a much more versatile weapon. Such a weapon could have two or more power settings: a high velocity setting for single-shot long-range precision shooting and a low velocity setting for shorter-range bursts of automatic fire.

The lower-velocity setting, of perhaps 3,200 ft/s using the same bullet currently used in the 5.56x45mm NATO would be approximate to the current performance of the cartridge. Such a setting allows automatic fire in a controllable manner, which is suitable for close quarter combat, or rapid aimed semiautomatic fire at targets at intermediate ranges.

If we take the recoil of the larger 7.62x51mm NATO cartridge as the maximum allowable recoil, we could take the 62-grain bullet from the 5.56mmx45mm NATO cartridge and accelerate it up to over twice the velocity of the larger cartridge, in the region of 6,600ft/s. In this scenario, the muzzle energy of the 5.56mm round increases by 4.5 times, and is 2.3 times as powerful as the larger 7.62mm round.

Such a bullet would be very flat-shooting, which minimizes the effects of wind and distance-estimating error, and would transfer very large quantities of energy to the target. This would make it a very effective sniper weapon and would represent the long-range precision shooting aspect of a variable-power weapon.

Much like the assault rifle combined the close-combat automatic-firing ability of the submachine gun with the accuracy and power of a full-power battlefield rifle, a variable-velocity railgun could combine the assault rifle with long-range sniping and/or armor-penetrating ability.

One problem with variable-velocity railguns is the rate of twist of the rifling. A rate of twist that works well at lower velocities might cause enough spin in higher velocities to literally tear the bullet apart from centripetal force as soon as the bullet leaves the barrel, and a rate that works well at high velocities might not sufficently stabilize a lower-velocity projectile. Most likely is that the difference between the high and low velocities will not be as great as described above, but even an increase of 1,000 ft/s at the muzzle results in nearly twice the kinetic energy at 500 yards and 60% more energy at 1,000 yards.

If the power source was small enough, it would also enable the solder to carry several times the ammunition than a conventional rifle cartridge, which is about 30% to 50% bullet (or payload) and 50% to 70% powder, primer, and brass case, by weight.

While it would be possible to reduce recoil by constructing a railgun which fires very lightweight projectiles at high velocity, such projectiles are extremely inefficient at wounding compared to larger, heavier projectiles at moderate velocity. While energy correlates with the amount of penetration in armor, this is due to the nature of the impact involved: metal on metal. Despite the prevalence of the "energy transfer" hypothesis, energy will only correlate with damage done in soft tissue if bullets enter with a large impact mass to speed ratio.

Additionally, the recoil of a hand-held railgun may be more of a problem than the recoil of a weapon with the same projectile momentum, but a lower velocity and heavier projectile. Since a high-velocity railgun would accelerate the projectile over a shorter time, the force exerted on the projectile (and hence, the recoil force exerted on the weapon) would be greater. For example, commercially-loaded .357 magnum (9 x 33 mm R) ammunition which uses 125 grain (8.1 gram) bullets at 1450 ft/s (442 m/s) is almost universally regarded as having more unpleasant recoil than ammunition with 158 grain (10.2 gram) bullets at 1235 ft/s (376 m/s). Despite the higher momentum of the latter load, the higher velocity of the former means that the bullet (and therefore the gun) accelerates faster, making the recoil more forceful. A railgun which fires a very light projectile at very high speeds would inevitably have an extremely high recoil force, though the impulse could be kept relatively low. A shock absorber built into the gun could compensate for this, but would add to the weight.