Quantitative Analysis of Shaking Time in a Martini

When making a martini, shaking with ice is a common technique to mix the ingredients. This has two easily measurable effects. The first is that the martini is cooled by the melting of ice. The second is that this melting dilutes the martini. The melting of ice seems to occur even if the gin and the ice are the same temperature initially.

Initial observation:
While gin is typically 80 proof (40% alcohol by volume), some gin is 94 proof (47% alcohol by volume). When making martinis with the higher proof gin, it has been noticed that the drink has a much harsher taste and mouth feel. It has been suggested that shaking with ice longer will melt more ice where therefore lower the concentration of alcohol in the drink, thereby reducing the harshness.

Will shaking a martini longer melt significantly more ice and dilute the drink further?  If so, how will this affect the temperature and flavor of the drink?

Since a water ethanol solution has a much lower freezing point than plain water and it has been observed that ice melt occurs even when the gin and ice start in thermal equilibrium, it is reasonable to assume that ethanol has an affect on ice similar to that of salt allowing it to melt at a lower temperature. Since melting requires heat, this chemical melting is expected to be endothermic resulting in a lower temperature. If this process is allowed to continue for  longer time, a lower temperature and more melting is expected.

Equipment & Materials:


  1. Record masses of empty cocktail shaker, graduated cylinder and cocktail glass.
  2. Shake 6 fl. oz. of cold tap water in the shaker to get the shaker to a consistent starting temperature.
  3. Measure 3 fl. oz. of gin, plus 0.5 fl. oz. of dry vermouth, and mass graduated cylinder with liquid.
  4. Take temperature of liquid using instant read thermometer.
  5. Place 4 ice cubes into shaker and remass shaker.
  6. Take temperature of ice using infrared thermometer.
  7. Add liquid to shaker and record mass of shaker with ice and liquid.
  8. Start 30 second timer, and shake vigorously.
  9. Strain liquid into cocktail glass.
  10. Measure temperature of liquid in cocktail glass using instant read thermometer.
  11. Measure temperature of ice with infrared thermometer.
  12. Remass filled cocktail glass.
  13. Remass shaker with ice still in it.
  14. Repeat steps 1 through 13 but shake for 60 seconds in step 8.
  15. You should now have two Classic 1950’s Martinis (as described in Vintage Spirits and Forgotten Cocktails), shaken not stirred.  For best results add a dash of orange bitters to each, stir, and enjoy responsibly.


30 seconds 60 seconds
Pre-shake water (before) (ºF) 46.7 46.6
Pre-shake water (after) (ºF) 49.8 48.2
Before Shaking
Graduated Cylinder (g) 109.9 110.4
Shaker (g) 300.2 300.0
Glass (g) 188 184.6
Shaker with Ice (g) 412.1 403.9
Ice Temp (ºF) 2.2 4.2
Graduated Cylinder with liquid (g) 212.2 212.3
Liquid Temp (ºF) 16.9 26.7
Shaker with ice and liquid (g) 513.5 505.5
 After Shaking
Liquid Temp (ºF) 13.5 19.9
Glass plus liquid (g) 306.0 314.7
shaker with remaining Ice (g) 394.4 374.9


30 seconds 60 seconds
Ice before (g) 111.9 103.9
Liquid before (g) 102.3 101.9
Temp change (ºF) -3.4 -6.8
Liquid after (g) 119.1 130.6
Ice after (g) 94.2 74.9
Liquid change (g) 16.8 28.7
Ice change (g) -17.7 -29

Based on the measurements taken, shaking for 60 seconds instead of 30 seconds does in fact cause more melting of ice ~29g instead of ~17g.  The temperature change was also greater, 6.8ºF for 60 seconds as opposed to 3.4ºF in the 30 second trial.  The temperature change however is somewhat suspect as the gin was not returned to the freezer during the first trial and warmed up significantly.  As a result the different in temperature is bigger, but the final temperature is actually still higher.  This temperature difference could also account for some, if not all, of the additional melting.

In terms of qualitative analysis, the 30 second martini was preferred.

Surprise Results:
When shaking the water before the 60 second trial, the water mixed with the martini residue from the 30 second trial and became very foamy and was actually effervescent when the shaker was opened to take a temperature reading.

The results indicate that shaking for an additional 30 seconds does cause more ice to be melted.  The temperature drop was also greater when shaken longer. However these results are slightly tainted by the failure to return the gin to the freezer between trials.

Future Questions:
This study was carried out with 80 proof gin as there was insufficient 94 proof gin was available for two trials.  If this were repeated with 94 proof gin, would the results be consistent?  Would the difference in melting be greater?  Would the taste preference switch from 30 seconds to 60 seconds?

One thought on “Quantitative Analysis of Shaking Time in a Martini

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s