[Ed: We published this post challenging the conventional wisdom on longer barrel ballistics from Josh back in October of 2013, before he became a regular around here. It provoked much discussion at the time and probably will again the second time around.]

Abstract: This is an independent scientific study that has been conducted in western Michigan. This study addresses the misunderstanding of the concepts related to barrel length, muzzle velocity, and accuracy in a rifle . . .

Elements of the Study: This study was conducted with a set of standards that do not necessarily correspond to all manner of firearms. The combination of weapon and ammunition used for this study was carefully determined and analyzed for the best results. This study was conducted with what the author and fellow researchers determined to be the most precise materials and methods available gathered from expert input and other existing studies.

The platform used for this is a Shilen match barrel which began at 26 inches in length and ended at 13.5 inches. The chamber is of standard SAAMI specification in .308 Winchester and the barrel features a 1:10 right hand twist. The ammunition used for this test is of several types, all of which are of corresponding lot numbers. At each range, handloads were used to seek out advantages given the barrel length by modifying the bullet and powder. This data is included gratis and represents the abilities of the weapon system when tuned ammunition is available.

For this test, the barrel was attached to a Savage short action target receiver in a Scally Hill Systems MK4 Mod7 folding chassis. This test measured all three variables at the same time in the most similar conditions available. Testing was conducted at Southkent Sportsman’s Club in Dorr, Michigan and Chick-Owa Sportsman’s Club in Zeeland, Michigan. Firing was conducted at a distance minimum of 100 yards and a maximum of 540 yards. Informal ‘field’ shooting was conducted on private land at safe targets out to a distance of 900 yards, accurately measured by satellite using Google Earth.

Ambient conditions were on average 70-75 degrees Fahrenheit with 40-50% humidity at an elevation average of 670 feet.  Shooting was conducted with a 16x SWFA SS optic, a piece well noted for its durability and ruggedness. Velocities were obtained using a chronograph and extrapolation of shooting results. Group size was measured with a micrometer. Five shot groups were used to measure accuracy. Firing was conducted on standard IPSC silhouette targets at all ranges.

The truth about barrel length accuracy velociity

This study does not aim to look at terminal effects, rates of drop and drift, combat effectiveness, ethical viewpoints, or legal/political issues.


This section is included here as a semi-abstract to address commonly held beliefs regarding barrel length, muzzle velocity, and accuracy. These results are backed by the data collected below.

Explaining Barrel Length:

Belief: a long barrel is required for accuracy when shooting at long distance.

Fact: In no part of our testing was barrel length a determining factor in accuracy. At a distance of 100-540 yards, there was no discernible difference in accuracy between various barrel lengths. This performance translated over to unknown distance shooting with all barrel lengths at ranges out to 900 yards. At no point in the testing was a short barrel a hindrance once marksmanship fundamentals were observed and proper flight data was applied.

Explaining Velocity:

Belief: Now that we know that accuracy is pretty much the same, short-barreled rifles lose too much velocity be effective at long ranges.

Fact: This is a double-edged sword. The 13.5-inch length could propel a 168 grain Hornady TAP round at an average velocity of 2390 fps, which is hardly slow. That is only a decrease of around 315 fps from the 26 inch length (25.2 fps/in), and vindicates many researchers who pioneered velocity discussions. There was no noticeable critical difference in accuracy at any range. There is a downside to longer ranges and reduced velocities, that being increased susceptibility to wind as range increases. Increased drift is not the end of the world, though, and if measured properly, can be overcome with ease.

What is more is the differences in velocity across loads and barrel lengths. The issue with barrel length and velocity was also interesting in that, across all bullet weights, the extreme variation is only 31% (110 VMAX @3202 and 208 AMAX @2215). In the most accurate load, the 168gr HPBT handload, the velocity difference between longest and shortest was only slightly more than 15%. The round with the least variation between barrel lengths was the 175gr Federal Gold Medal Match with slightly less than 8% variation.

Explaining accuracy in a practical sense:

Belief: “The time I put five shots into a clover leaf is the time I did everything right.”

Fact: This is the greatest misunderstanding in the world of accuracy and shooting. In our testing, no matter the ammunition used, the weapon showed that there was a natural fluctuation in regard to group size and point of impact. This has been determined by other studies as well, even those using ‘rail guns’ and heavy benchrest rifles. Accuracy, at least in our testing, was determined to be more akin to a ‘cone’ than a grid in that the accuracy of the rifle had an average maximum radial spread of .765 MOA over all barrel lengths.

In layman’s terms, this means that the barrel could fire an indefinite number of rounds into a circle with an average diameter of 1.53 MOA, which is not all that impressive. However, it must be understood that accuracy does not work like traditional manuals dictate. As an example, a man takes his new rifle to the range. He sets up his targets and fires several five shot groups. His groups are respectable by most standards, with most clustering at around .75 MOA. He sets his zero and continues to fire.

Here is the important part: he fires another group and gets a ‘flier’ one MOA low and left. He discards it and continues, discarding all the fliers he gets. Now it gets hard for him to figure out. He shoots five shots and notices that he gets a .25 MOA group, but .8 MOA low and .45 MOA right. This is a great group, and he scratches his head and adjusts his scope to it. He shoots again, but prints a wide group measuring 1.2 MOA across, but now shifted off his zero. He assumes that he has run his luck out, packs up, and goes home.

What has happened here has happened to many people. What our friend did not realize was that his gun was never zeroed at all. The tight cluster he got was not the time he did it right, it was a statistical possibility that comes from firing. In reality, the man had a rifle that was not shooting .75 MOA, but rather he was printing groups and ignoring his most important ally, his fliers. These are critical to rifle accuracy and are not mistakes.

Statistically speaking, the rifle he has may actually fire a maximum group size of 2 MOA at 100 yards, which sounds terrible, but really isn’t. The vast majority of his rounds will probably impact at a radius of around .5 MOA of his true zero, or even less if he has a good combination. What he did not understand was that there is nothing wrong with a rifle that may throw a round out even 1 MOA or more, it’s all within the statistical level of accuracy that the rifle is capable of.

The results of this study were very telling. Overall, as demonstrated in the accuracy charts, the shortest barrel length provided the most consistent accuracy across the board and the longest length proved to be the least accurate with the same loads.

The data also shows that the so-called “MOA” a rifle can shoot changes with distance. The groups at 100 yards show very good, often benchrest-grade accuracy. And then at extended ranges, they show a natural increase in group size.

Across the board, all the loads tested across all barrel lengths showed this. Across all loads and lengths, the average at 900 yards was .765 radial MOA or 1.53 MOA. Compare this to the 100-yard average of .206 radial MOA or 0.413 MOA. That’s right, the average across all lengths and loads yields sub half-MOA at 100 yards and just over 1.5 MOA at 900 yards.


This test obliterated what was previously thought to be fact. Not only was it determined that short-barreled rifles are easily as accurate a those with longer barrels, but we also discovered what we see as a key to viewing accuracy in a practical sense.

In an age of misinformation, hard fact can be hard to come by. The internet is full of armchair know-it-alls and trolls a-plenty, but for the most part, these can be ignored. Mental preconceptions of the researched concepts are still deeply entrenched in a more or less Napoleonic era of the theory of arms.

Most of what is commonly argued about small arms is false and based on opinion. A quick look online reveals hundreds of arguments on topics like 9mm vs. .40 S&W vs. 45 ACP or AR-15 vs. AK-47, none of which are based on fact or on the need of the individual in their realistic circumstances.

If anything is to be learned from bullet selection, it is that match-quality bullets have a distinct edge in accuracy over military and hunting bullets. The match bullets tested produced significantly greater accuracy than their military or hunting-type counterparts.

This study is not aimed at the promotion of using any particular barrel length, brand of bullet or load. The reader must look at their own situation and determine what the most valuable features are in a rifle.














Josh Wayner is a senior at Grand Valley State University studying Applied Research Science in the Interdisciplinary Studies Major. He has been a competitive shooter for nearly ten years and has eleven CMP medals from Camp Perry. Josh is also the owner of Scally Hill Systems and is constantly developing and testing new things. Long range shooting has been his passion for many years and he continues to push the envelope in modern rifle design. 


  1. Can we transpose the conclusions onto pistols? A snubby revovler is just as accurate as a 6in barrel target pistol of the same caliber?

    • Snubbies are accurate, but their long, heavy trigger, minuscule sights and short sight radius make them hard to shoot accurately

      That is the difference between inherent accuracy and practical accuracy

    • I consider myself an average shooter- but regularly shoot tighter groups at the range than those with full sized semi autos (where most targets look more like someone is patterning a shotgun than excersising skill with a handgun). Like most weapons, it comes down to shooter skill and familiarity with the weapon. During my CPL training an officer was making perfect cloverleafs, which none of the others present could duplicate even using the officer’s weapon.

      • Was at the range once a year or so ago (maybe more). the next three lanes were police and sheriff (3 to a lane). Leaving, I was able to actually see their targets. Wouldn’t want to be in the same town when they had to unholster against a suspect.

        • It’s very true that Police and Sheriff’s aren’t able to shoot well. I shoot with a group that helps a Sheriff team practice and learn. They are some great people to know, but many departments don’t make shooting a priority. It’s a real shame considering the lifes at risk, including their own. I’m happy volunteering my time to this group to help out that department. Maybe more departments should work with civilian shooters to improve their skill set, and budgets adjusted to allow the LEO community to practice with ammo supplied by the department. Most don’t and that is the main problem in my opinion.

        • “Maybe more departments should work with civilian shooters to improve their skill set, and budgets adjusted to allow the LEO community to practice with ammo supplied by the department.”

          More officially sanctioned and funded training might be politically counter-productive. While the anti-gun mob is convinced regular people are not trained enough to be trusted with guns, more training for LE might lead to protests (you have to be anti-cop to be a good Leftist) that municipalities are training cops to be eager killers. Same if the public discovered a cop shooting one or two extra boxes of ammo on their own time.

    • Yes… and no.

      First, we’d have to use optical sights on both the 6″ target pistol and the snubby. Otherwise, the snubby will suffer much more from the decrease sight radius than from ballistic effects.

      Second, snubbies are of a length where the muzzle velocity is taking a drastic turn downhill:

      Look at the pistol MV’s for various barrel lengths at the following site:


      See how the MV’s start to go downhill rapidly with barrels under 4″ in length? When you stop to think about percentage loss of potential velocity by going from a 6″ to 2″ barrel, it is substantial – in nearly every handgun cartridge.

      At these substantially lower velocities, you’re going to see the effects of wind magnified on the slower rounds that take longer to reach the target at 25 yards (or further). Keep in mind that most handgun bullets have terrible Bc’s, and from a snubby, it not only starts out slow, it gets slower pretty quickly compared to a rifle round.

      Add in a little wind and you’ll see your groups start to come apart at 25 to 50 yards with a snubby based on nothing more than the low velocity and low Bc.

      • The 3″ revolver BBTI used for the .357 magnum must have had a huge cylinder gap. Also, they don’t have any full power loads. You should get 475-500ft/lbs out of the regular factory stuff and ~600ft/lbs out of the Buffalo Bore, Double Tap, etc. (non +p) loads.

    • They tested mechanical accuracy of a barrel/load. Snubbies shoot horribly because of their long trigger pulls and short sight radius.

      Put it in a random rest against a Glock if you want to compare mechanical accuracy.

  2. No discussion of the rounds going transonic?

    It’s been my understanding that this is the reason many .30 cals are no longer considered ideal for long range work. The data I *think* seems to indicate a large change at the 900yd range for most of the rounds, and if I have the BC close on the 155gr bullet, its going around 1100FPS by the time it gets there with the 13.5″ barrel. And I’m willing to bet the boat tails do better at that threshold than the shorter, flat based 110gr bullets.

  3. Sorry, but if you’re testing barrels, you need to take everything else out of the setup (aka Scientific Method 101). The barrel (and action, of course) should have been clamped into a completely mechanical jig and never moved after that, and the rounds fired mechanically. Since that wasn’t done, this was a test of the barrel+human combination, and no conclusions can be drawn about barrels alone.

    • True, it’s a test of the barrel-human combo (not to mention a hundred other factors that can alter the path of a bullet). But if the human and other factors stay the same and the barrel is changed, then any statistically significant differences can be attributed to the barrel.

      Scientific method 101.

  4. Often, when i shoot a flier from a rifle, I *know* it was a flier, and know i did something wrong. so ‘count your fliers’ is not always going to be a legitimate approach.

    • Agreed, I often ‘feel the flyer and can call it without looking at the target. I love the concepts in the OP, I hate that I’m agreeing with a second criticism of it.

  5. Personally I’ve always liked 18-20″ ba rrels since it seems like the velocity loss in 16″ barr els aren’t worth it. This seems to validate that. By my math the average of the 18″ ba rrels was about 95% that of the 26″ ones whereas the 16″ averaged about 97% of the 18″.

  6. Traditional thinking was that longer .308 barrels were faster and more accurate.
    20 inch AR 10 stainless bull barrels were the “target” barrel of choice.
    Modern thinking has changed to believe that 18 and even 16 inch .308 barrels are just as accurate, but give slower velocity.
    This test seems to confirm that general idea.

  7. There’s not a lot here that changes what accuracy gunsmiths have known for a couple decades now. Shorter barrels are stiffer. That’s just basic mechanical engineering: Young’s Modulus is coming into play. If you want to close up groups while keeping a barrel long (and thin), you can use a barrel tuner, or something like the BOSS tunable muzzle brake. As you move the mass further out/in on the barrel, you can find a point where the vibration deflection is at a minimum as the bullet leaves the barrel, and you get tighter groups. You see tuners used frequently on .22LR target rifles, where you don’t get to control the ammo variables.

    I’ve shortened several customers’ barrels from 24″ to 18″ or 20″, and they’ve seen no degradation of precision downrange. I always tell people how much it will cost them in velocity – 25 to 45 fps per inch of barrel removed (this depends on the cartridge and barrel maker), down to about 17″, so they know what they’re getting into. On a .308 out to 800 yards? Hack off as much as you want, down to legal length. Out to 1,000 yards? Well now, those 200 fps you chopped off might start to become important with some bullets/loads, because the pill might go sub-sonic before 1,000 yards.

    Several times, I have recommended that folks here pay attention to what the benchrest folks have learned since the 1960’s, but y’all seem bound and determined to re-discover knowledge that is fairly well known in benchrest circles. What Josh is saying here about statistics (in a rather non-mathematical way) is true. One of my rants about rifle group size stats is the standard gunsmith rant about 3-shot groups. If you grind through the stats and sampling math, you find out that you can obtain 3-shot groups from a factory rack-grade 2MOA rifle that will look damn impressive – on a fairly regular basis. The flip side, however, is that you’ll also obtain a bunch of 3-shot groups that don’t impress – on a fairly regular basis. It’s just statistical dispersion you’re seeing in both cases. What you won’t get from that rifle (often, if at all) is a five or 10 shot group without fliers.

    Getting a rife to print a tight 5-shot group? OK, now you’re getting somewhere.

    Getting a rifle to print a 10-shot group under 0.25″ at 100 yards? Back up the Brinks truck, you’re going to pay to obtain that. The current record-holding benchrest group is 0.0077″ from five shots at 100 yards, set in 2013. This broke the 0.009″ 5-shot group from 45-odd years ago. NB what I just said there: 45 years ago, benchrest shooter & gunsmith Mac McMillan made and shot a gun that laid down a group under 0.010″. So it can be done…

    Now, for something else you should know: barrels from different barrel makers have different muzzle velocities – with the same bullet, same load, same barrel length, everything. There are some premium barrels that are “faster” than others – sometimes by more than 150fps in a .308 or similar cartridge. Shilen, in my observation, has not been one of the “fast” barrels.

    • Correct me if I’m wrong, but doesn’t a slower rate of twist result in higher velocities? Also I’ve heard that Marlin’s micro-groove rifling produces higher velocities, although that’s not going to do much for anything hotter than a .30-30.

      • Yes, a slower twist can increase MV’s.

        However, longer bullets (longer bearing surface specifically) need a tighter twist to stabilize properly. Higher-Bc pills will typically be longer, because a longer bullet has a higher sectional density, and this leads to higher Bc’s (for a given form of bullet).

        This is why the “gain twist” rifling idea keeps coming back into barrel making again and again and again and again… gunsmith after gunsmith has chased the idea of gain-twist rifling (where the rifling starts out with a “slow” twist and the rate of twist tightens as you get nearer the muzzle) as a way to increase velocity, decrease gas blow-by, increase stability, etc. Harry Pope built his own rifling machine, and played around with gain-twist rifling quite a bit. Here’s a picture of Pope’s rifling setup:


        You see the compound curve on that sine bar? There you go. Gain twist.

        • Yes, a longer twist rate will limit your high SD choices. But if you want say, a .308 for deer hun ting out to 300 yards or so and only intend to use 150 or 165gr. pills you’ll get higher velocities from a 1 in 12 twist over a 1 in 10, as I understood it. Not sure if it’s 5fps or 100fps though.

          I would think that that gain twist rifling would waste anything it gained because instead of the bu llet deformation occurring totally in the first inch (where pressures are at their highest) it would be continuously deforming all the way down the ba rrel. I’d think that would put a lot more heat into the barr el and bu llet as well.

  8. The data indicates a trend that the “shorter” barrels offer greater precision.
    A more precise explanation is that barrels with a higher stiffness to mass ratio result in greater precision. This is why fluted barrels are often recognized as higher precision – because it increases the stiffness to mass ratio.
    Really great analysis and discussion on the subject:
    So, the questions are:
    What is the maximum terminal effect required and at what range?
    What is the shortest barrel possible to deliver this and in what chambering also taking into consideration the statistical effects of drop, and wind drift on hit probability at range?
    Basically a high velocity VLD projectile increases hit probability at all ranges.
    So it really comes down to optimizing all these factors based on the maximum weight of the rifle and ammunition you are willing to carry, and the round count required for the mission.

    • “A more precise explanation is that barrels with a higher stiffness to mass ratio result in greater precision. “

      Exactly. That’s what Young’s Modulus is all about: Stiffness of a material. It is an intrinsic property of a material; ie, hardened steel will have almost the same Young’s Modulus as soft steel. The YM of the fundamental material is what is determining the stiffness, and in this case, it is Iron (Fe).

      Here’s a table of some properties:


      NB how steel, iron, wrought iron, stainless steel have YM’s in the same ballpark.

      If we want a stiffer rifle barrel, we need to find a different material to make them. Carbides are much stiffer, for example. This is why machinists use carbide boring bars with carbide inserts when they need a “long” boring bar. The carbide bar is stiffer than any steel boring bar can be…

  9. There is no way in hell a 13 and a half inch barreled .308 will out shoot a 6.5 Creedmoor with a 500 yard barrel at 540 yards, no way

  10. For most people who are actually knowledgeable about guns, this has been known for a long time.

    I would argue with the author about how the increased wind drift caused by the reduced velocity of shorter barrels is easily compensated for. When you are shooting at long range the two biggest challenges are precise range estimation and wind correction. Even shooting targets during competitions with targets at known ranges like F Class or High Power, the winners are usually the ones who make the best wind calls. So, in a no wind situation at long range with targets at known distances a short barrel will have the same or better mechanical accuracy compared to a longer barrel. But, in real world conditions will probably fall short. There is a reason 6mm and 6.5mm cartridges are dominant in the Steel Safari / Sniper type competitions where targets are shot at long range at unknown ranges and varying elevations. Their high BC projectiles help make up for small errors in wind / distance / angle calls by shooting flatter and drifting less than traditional cartridges like .308.

    What I think this study does is show that for any cartridge their is a point of diminishing returns on barrel length. For the .308 used in this example that point is anywhere between 18-22” depending on what you are going to do with the rifle. Another example is 5.56 where anything more than 20” is pointless and 18” doesn’t cost you very much at ranges less than 500 yards.

  11. This study seems to confirm Jim Owens writings of cone dispersions in Sight Alignment Trigger Contol and the Big Lie. The book is at least 30 years old.

  12. Factory ammunition is not consistent. There may be a difference of say 100 FPS from fastest to slowest round in a box of 20. That difference is enough to affect the POI even if the rifle is actually zeroed and bench rested.
    The only way to guarantee accuracy is handloaded ammunition. Same sized, trimmed and weight of the brass, projectiles hand selected for identical weight, same for powder charge, seating depth and crimp if used.
    I use Hornady 22Cal .224″ 75gr. BTHP, Lake City 5.56 NATO brass, 21.6gr. H335 and OAL 2.250″ in my suppressed 16″ 1:7 twist 5.56 NATO barrel, bench rest with bipod at 50 yards (all the room I have in my backyard) and can have 5 shots touching each other.
    P.S. I had a 100 yard target range, but Hurricane Florence took care of that by downing some huge trees in the woods blocking it off. Too big for my chain saw to handle.

  13. Doesn’t the case capacity and power burn rate come into play here? A 5.56..223 case has very little capacity compared to a .30 06 not to mention magnum cases. A .308 has a relatively smaller capacity that the .30 06. The more quantity and slower burning powder should take advantage of a longer barrel if you have more propellant to continue accelerating the projectile down the barrel. I read an article some years ago about the .30 06 being used with a 30″ barrel just to eke out all possible velocity vs the .308.

  14. “If anything is to be learned from bullet selection, it is that match-quality bullets have a distinct edge in accuracy over military and hunting bullets.”

    Well no shit. This info is not new and has been known for a very long time.

  15. The author seems unfamiliar with the work of Brian Litz. All this is well known, and well documented.

    Like group size expansion– it’s effectively linear with time of flight, not distance. No surprise there.


Please enter your comment!
Please enter your name here