So I was stirring a martini this evening, and it occurred to me that I should really just go ahead and see if shaking or stirring a cocktail adds more ice melt. Ice melt is the amount of water added due to ice cubes melting during the mixing process. People often think that shaking will add less due to the fact that stirring takes so much longer. However, that’s theoretically not the case, at least according to bar tenders I’ve talked to and read posts from. I tend to think the bartenders are right, and certainly the texture is reason enough to make your martini stirred, but I decided to do a quick and dirty experiment to see if I could really see a difference. Here’s what I did:

  1. Made a martini. I’d been wanting to go back to the Junipero gin I picked up a few months ago, so I did. Still not a favorite of mine, but it makes a fun martini that certainly lives up to the “juniper” part of the name.
  2. Poured some water in a mixing bowl. The water was approximately room temperature, and then I let it sit for a while. I didn’t want the temperature to change between the two mixings.
    Bowl & Measuring Cup
  3. Ensured my shaker was dry from previous use…
  4. Carefully measured out 1/3 a cup of the room temperature water. This picture is a little blurry as it’s somewhat difficult to get a good shot of a clear liquid in a glass container.
    Water Pre-Shaken
  5. Added 4 ice cubes. For both the shaken and stirred I tried to use similar sized cubes, which meant a bit of eyeballing as well as removing areas where the water had sat over the dividing section and froze. Not particularly exact science here, but hopefully close enough.
  6. Shook the mixture. I tend to want my drinks cold, so I shook it about 70 times. That’s more than normal, but I was exaggerating things a bit to hopefully produce a visible result.
  7. Dried the measure cup and poured the water back in. You can see that a fair amount of water was added:
    Photo 1: Shaken 1
    Photo 2: Shaken 2
  8. Then I dumped out everything, rinsed and let it sit for a while to return to room temperature. While this was going on, I dried the shaker.
  9. Repeat the process, but stirred 70 times. This took considerably longer, and I took my time since stirring should take longer. Here is the water before I added it to the ice in the shaker:
    Water Pre-Stirred
    It looked like possibly a hair more than what I used for shaken, but very close.
  10. When I measured the results, they were similar, but the stirred does seem to have less ice melt.
    Photo 1: Stirred 1
    Photo 2: Stirred 2

The main conclusion is that I need more precise measure instruments. That being said, it does look to me like stirring the drink adds less ice melt. This backs up the general consensus by those wiser and more experienced than I. Still, it’s nice to have confirmation. Even if it’s not terribly dramatic, the results are enough to satisfy my curiosity. In the end, the Junipero martini was tasty, so I consider the experiment a success!


  1. You are so nerdy on so many levels! I love it. The results look almost identical to me. What’s really impressive is the amount of ice melt involved either way. I had no idea it was so much.

    Damn, I want a martini, and it’s only 9:43 a.m. Thanks, Ardenstone.

  2. Ardenstone

    The results were close, and unfortunately I didn’t set up the photos so that the camera was the same distance away, etc. I should have marked the measuring cup after the first one with a bit of tape or something, but this was an off-the-cuff activity. The ice melt really was a lot, although I did also shake and stir for more than one typically would.

    Mmmm, breakfast martini…

  3. ive always wondered what the difference was, so this was actually pretty helpful…
    that is until the very end. im a little confused because next to the number 10 it says that shaken had less ice melt, but then in the final word it said that stirred had less ice melt.
    so im curious as to which one was the mistake.

  4. I cleared up the wording it a bit. This experiment really was a wash, so I wouldn’t take it as very conclusive. I’ll need to try it again some day!

  5. What about the issue that liquor only cocktails should never be shaken and cocktails with fruit juice and liquor should always be shaken to make sure the 2 or more ingredients are well mixed? I have heard from bartender friends that are purists that liqour only cocktails mix better and hold up better when only stirred. Your comments please. thanks, Patrick

  6. Patrick,
    I actually tend to follow that rule, although I haven’t heard the reasons you describe. Anything clear, I stir. Anything opaque (including anything with cream or fruit juices in it), I shake. Primarily this is a cosmetic decision: translucent drinks aren’t as translucent after you shake them. A shaken martini just doesn’t look as nice as a stirred one. You could also stir the opaque drinks, but it is faster and easier to shake them. Also, most of the opaque drinks I serve are just fine with a little foam or bubbles along the rim, an effect which I find less pleasing in a translucent cocktail.