Yes, the 9mm Ammo Quest continues. I may occasionally get distracted with shotgun slugs or AR15 tests or the new exotic .380 ammo or whatever, but the main grind is to get through these 9mm tests and find what works best and what won’t perform in a 3″-barrel pistol. I tested the Speer Gold Dot 115-grain a while back, and it was a really rather good performer. The heavier Gold Dots hadn’t fared as well from the 3″ barrel, but the light 115-grain bullets did much better. After that test was released . . .

several people asked for a test of the “+P” version. Well, Speer doesn’t make a 115+P… but, apparently, other manufacturers do. Specifically, Underwood and DoubleTap both offer 115+P defensive loads that, while they don’t say “Gold Dot” on the box, they sure do walk, talk, and quack like a Gold Dot. It appears to be a marketing restriction; apparently Speer sells their bullets to other manufacturers for them to use, but doesn’t authorize those manufacturers to use the “Gold Dot” trademarked name. Accordingly, Underwood sells theirs as a “115gr. GD J.H.P.”, and DoubleTap offers theirs as a “115gr Bonded Defense JHP”.

But make no mistake, these are Gold Dots. I wasn’t sure when I started the tests, but one look at the expanded bullets will erase all doubts — these are the genuine Gold Dot bullet.

So both Underwood and Speer offer a 115+P version of the Gold Dot. How do they compare to each other? Is one better than the other? I decided to do a combined test, to answer those questions. And the answer is….

… they’re Gold Dots, loaded to a higher velocity. They perform extremely similarly to each other, in terms of expansion and penetration. The DoubleTap actually clocked in at an average of about 38 fps faster than the Underwood, and both of them showed a notable increase over the Speer standard-pressure load. The Speers, in my test, delivered 1123 fps; the DoubleTaps averaged 1245 fps and the Underwoods averaged 1207 fps. So we’re looking at a speed increase of around 84 to 122 fps faster by going to these +P loads vs. the standard pressure loads. Recoil increases commensurately, as you’d expect.

What you may not expect is what happens with the penetration, though. Both of these +P loadings showed quite a bit LESS penetration through the bare ballistic gelatin than the Speer standard pressure version! Whereas the standard pressure version penetrated to 12.50 to 13.00″, the +P bullets came in at between 9.50″ and 11.00″. That’s 2 to 3 inches shorter penetration. Why? Because these faster bullets used that additional energy to expand to simply gigantic proportions.

The bare gel bullets are massive starfish, as opposed to the standard pressure loadings, which were textbook typical Gold Dots. The standard pressure bullets expanded to an average diameter of .554″, with a maximum size of .646″. That’s pretty big, considering the bullet starts out at .355″ in the first place, so at .646″ the standard pressure bullet expands (at a maximum) to be 82% larger. But the Underwood and DoubleTap speeds push that expansion to behemoth proportions (well, a behemoth of a little 115-grain bullet). The Underwoods grew to a maximum expansion of .787″, and the faster speeds of the DoubleTap had them growing even larger, to .808″. That’s 127% larger than the original diameter — it’s extreme expansion.

But, the more a bullet expands, the less it can penetrate. Expansion acts like a “parachute”, increasing the drag on the bullet as it travels through the gel (or flesh or water or whatever you’re shooting it into). The bigger the parachute, the quicker it slows down. The initial “smack” and the initial temporary cavity are bigger with these faster bullets, but the overall penetration (in the bare gel) is quite a bit shorter than the standard pressure versions.

Surprisingly, in the denim, the faster +P bullets actually expanded a tiny bit less on average than the standard pressure loads did (about .455″ for the +P versions, vs. about .472″ for the standard pressure bullets). I don’t have a good explanation for you on that; the overall size and penetration was generally about comparable between the +P versions and the standard pressure. Penetration was fantastic; the DoubleTaps averaged 15.80″ and the Underwoods reached 16.30″ on average.

Through denim, all three of these Gold Dot variants performed great. Through the bare gel, I would have to say that the standard pressure performed better, from the 3″-barrel pistol. The +P versions weren’t “bad” through the bare gel; they averaged 10.40″ for the DoubleTap and 11.00″ for the Underwood. That’s probably okay, but not ideal; the established standards for bullet performance call for a minimum of 12.00″ of penetration through ballistic gel. An average of 10.40″ is a little short of that, but given how excellent the denim performance was, and that in actual autopsies the recovered bullets generally look more like the bullets recovered from denim-covered gel testing, I’d say you’re probably just fine with the +P versions as offered by these companies.

The question to consider then at the end of the experiment is, is it worth it? What do you really gain by going to the +P version, and is it enough to make up for what you lose? In my opinion, it’s probably not. The higher velocity would give you better performance at long range. But a 3″ pistol isn’t meant for long range engagements anyway, so that’s a moot point. The standard pressure gives plenty of penetration and good expansion. The +P versions give better expansion but even less penetration. And, using +P means more recoil, which may slow down your follow-up shots and means more wear and tear on the gun.

If it were me spending my money, I’d go for the standard pressure Gold Dots and skip the 115+P. If you really want a +P Gold Dot, I got superb results from the Speer 124+P Short Barrel ammo, and I would expect comparable performance from either Underwood’s or DoubleTap’s 124+P versions as well.

Recommended For You

25 Responses to Ammo Quest: 115+P Gold Dots: Underwood vs. DoubleTap

    • Search around on the net and you’ll find the 50 round boxes (supposedly intended for law enforcement) for $25-30. Same with .40 and .45. Not quite Winchester white box, but cheap enough to throw a magazine downrange once in a while.

  1. I love this guy and his tests, BUT….it would be nice, like real nice, if he had a RANKING of the rounds he has tested for a quick reference. Out of the gazillions visitors to this site I am sure there are folks who just want to see the top 3 rounds for each of their guns (3″ 4″ and 5″).

    • He did that with the .380’s once he was done with testing; I imagine we’ll have the same once he’s done with 9mm.

      Just from memory, I believe Critical Defense 115 grain, HST (all weights), Ranger T 147, Winchester Defend 147, Winchester White Box 115, CorBon DPX 115, and Gold Dot 124 +P all passed.

      • It’s pretty easy to remember “all HSTs” and that’s probably why SGammo’s email alert for 147HST is something like 2000+ people.

  2. The only way, still, and again, for a 9mm to gain credibility, is to bump
    It up to +P or +P+ to get a desired result, with 16+1rds.
    Sorry too many criminals in prison, rap albums, and videos, brag about their scars from a
    9mm. The bashing will commence about caliber in 3,2,1, etc et al.

    Almost identical about the .40 rd.

    As a disclaimer, I own an old school FN 9, love it. Jus sayin.

    • Just saying what exactly? Did you miss the part where he stated that the Standard pressure was a better performer and that the +p’s actually under penetrated a little due to over expansion?

      • At this point, my brain pretty much instantly translates “Just saying” into “I have nothing useful to add”.

    • I proudly introduce to you the Federal HST in any flavor. 124 standard pressure, 124+p, 147 standard pressure, 147+p. All perfect through any barrel length and denim or no denim it makes no difference. HST in any other caliber also works perfectly.

      • Agree. I use 124 grain +p in my Nano for better cycling and standard pressure in longer barrels. I used to go with a heavier bullet until I saw that velocity is only slightly higher than a 45 ACP hydroshock. If I want mass I will use a 45.t

        • With the exception of my sigma I’ve been stepping away from +p loads in my handguns. I’m not convinced that they bring enough more to the table to justify expense and wear on the guns.

          NATO spec 124 grain 9mm fmj is loaded to +p specs and for some reason that’s the most accurate load in my sigma. Whenever I see a box, I buy it.

    • Your anecdotal evidence means nothing. When it comes to hard data, .45 ACP and .357 Magnum do not fare substantially better than 9mm. A trauma surgeon can’t tell the difference, and one couldn’t tell the difference from analyzing statistical data about gunfights. But better control of the weapon and more rounds on target do make a substantial difference in a gunfight. Common “duty-power” handgun rounds (up to .357 Magnum) and “stopping power” do not go together. Anyone who claims any round commonly carried (anything up to .357 Magnum in power) can be relied on for “one-shot stops” is an ignoramus. Stopping power is the province of long guns (and hunting handguns as heavy as lightweight rifles unsuitable for carry). Even with a common rifle or shotgun, a “one-shot stop” is never guaranteed.

  3. Hey STB, have you considered making a website to aggregate all your reviews in one place and organize them by caliber, barrel length etc (kinda like BBTI)? You have accumulated quite a lot of very useful data already, and it would be even more useful that way.

  4. Just a personal observation: It seems to me that the denim is much more relevant because most people are covered by a relatively tough layer of skin.

    • I think you miss the point of Gel-testing. Denim does not simulate skin. It’s to simulate the clothing. It is relevant because we humans wear clothing to protect the vulnerable skin.

      Denim is used because it’s “heavy duty” and consistent. 4 layers are used to simulate a worst case “heavy clothing” possibly penetrating at multiple points (eg, going through both sides of an arm wearing a jacket, prior to penetrating the main body of the target).

      Ballistics gel is then used to “simulate” all of the soft tissue of the body in format that gives consistent standards. (the body has multiple tissues all with relative densities all of which would react differently – however it becomes impossible to get repeatable results).

      MAC has a really good writeup on it:

      “neither are capable of simulating human flesh, only approximating it. What’s important to me for my testing is consistency in performance. – See more at:

      So it’s an approximation of a simulation in a consistent form to allow repeatable standards based testing.
      Also why it’s important to not look at these Gel-tests in a vacuum, but instead use them as a knowledge base & apply real world results to it. Massad Ayoob over on Glock Talk frequently is asked about recommending particular loads as well as which departments use a particular load, as well as how pleased they are with their results.

      But that’s nebulous anecdotal evidence. Alone it amounts to personal opinion – when combined with STB’s or somebody else’s scientific gel tests you have a basis for making a judgement about what will work best for you.

      • There are a couple of problems with gel testing. I think you missed my point on the first one. The gelatin is supposed to simulate relaxed muscle tissue. However, before a bullet can interact with muscle tissue it must first penetrate the skin. With bare gel there is no attempt to replicate this at all. The best way to replicate this would be to use a layer of thin, soft leather in front of the gel. But I’ve yet to see anyone gel test that way, so I speculate that the denim test closer replicates a human wearing normal clothing rather than heavy clothing, even though they refer to it as ‘heavy clothing’.

        The second problem is that the gelatin isn’t really much like muscle tissue at all. You can push your finger right into gel. Go get a chicken breast and try pushing your finger into it. It doesn’t matter much at supersonic speeds because the nature of the medium isn’t all that relevant. It’s the density that matters. If it’s properly calibrated a BB should penetrate 3″-4″, so my speculation here is that the last 3″ inches of penetration won’t happen. Maybe in lung tissue, but not muscle tissue.

        Shooting into undressed animal carcasses would be much more informative, but I guess the PITA folks would have a fit over that.

        • No bullet is supersonic in tissue. Tissue is dense solid mass where the speed of sound is over 5000fps. Even rifle bullets are traveling at subsonic velocities in tissue.

        • The speed of sound is almost 5000fps – in water. In air sound travels at roughly 1100fps, depending on temperature and atmospheric conditions. When someone refers to a bullet or pretty much any other object as being ‘supersonic’ they’re referring to the atmospheric speed of sound, not the aquatic or the speed in which sound travels through ballistics gel.

          BTW, this is why you can’t tell the direction of sound under water.

        • No shooting into animal carcasses would not give “better” data. There’s to many uncontrollable variables.

          1st – Live tissue and dead tissue respond differently.

          2nd – Animals, even animals of the same species may have variation in skin thickness. Especially as decay sets in.

          3rd – the bullet track; the bullet is going to take different paths through different tissues with different relative densities all of which will impact bullet performance.

          The point of the gel testing is to control conditions so that each test has Similar conditions. It’s not supposed to be a 1-1 comparison between the “gel” and a living being. It’s to allow you to compare bullet to bullet in a controlled setting.

          —The second problem is that the gelatin isn’t really much like muscle tissue at all. You can push your finger right into gel.— Yes, we all know this. and it’s not a problem. That’s not why it’s used. As above it’s a simulation to provide for controlled conditions. If you can’t control the variables, you can’t repeat the test, you have useless data & bad science. Differing samples of muscle tissue even from the same source could have differing relative densities. Especially once it’s a dead source and decay begins.

          STB’s tells us his methods – calibrated ballistics gel, denim & a 3 inch pocket pistol. You could take those methods, your 3 inch pocket pistol, the same brand of ammo, replicate these results which should be similar.

          You could not do that with a Cadaver test.

        • ‘There’s to many uncontrollable variables.’

          Exactly. And gel testing simulates none of them. You could have a round that works great in bare gel but would plaster itself to the first rib it hits. Of course, shooting one bullet into a half rotted carcass wouldn’t tell you much, but a comprehensive study shooting dozens of rounds into different sized fresh carcasses would tell you a lot more than the gel.

          Personally I’m just taking a wild guess at how gel relates to tissue, but I prefer a round that shoots close to the 18″ maximum through denim. I think the denim somewhat replicates the skin. I think that at some point the bullet will be too slow to penetrate further in flesh but will continue in gel. I read somewhere that it takes about 250fps for an expanded hollow point to exit through the skin. I have no idea how accurate that is, but it does make me less concerned about over penetration, seeing as how the expanded bullet will have to penetrate the skin once to leave the body and another time to enter another. But like I said, I’m just guessing.

  5. Cool test. I’m loaded up with Federal HST 124 grain +P in me G23 w/ 9mm conversion and my Sig. It’s my favorite 9mm load.

    The G27 with Ranger is still on my hip, so it’s time to switch up carry for the weekend.

  6. *Just a note*
    .808 is WAY more than 127% bigger than .355. We’re talking about AREA, not diameter: doubling diameter QUADRUPLES area. A .808 bullet actually has a frontal area 5.18 times bigger than a .355 one!

    • Yes, I was referring to diameter, not area. In terms of area, it would be 5x if it was a soft lead projectile that mushrooms; however, these are bullets that use petals that splay out, so there ends up being “empty” space between the petals. So it’s not equivalent to a circle that’s .808″ in diameter, you have to average the maximum and minimum expansion figures together to get the real increase in surface area. That’s why I always quote the two figures separately — maximum expanded diameter, and average expanded diameter. On the DoubleTaps the average diameter was .630″ in the bare gel.

      The increase in diameter from unexpanded (.355″) to average expanded (.630″) is very large, it’s 1.77x. However, while the increase in surface area isn’t quite the 5x you had calculated, the increase in surface area is still huge — it’s 3.148x bigger than an unexpanded bullet. That’s tremendous expansion!

  7. I’ve found through my own personal evaluation that the Golden Sabers (124 grain standard & +P) and Critical Defense/Duty (124grain/135 grain standard and +P).
    These are through a Sig P938, and are just the best for what I like to see.

Leave a Reply

Your email address will not be published. Required fields are marked *