Post Format

The Ultimate Thermos Test: You’ve Never Heard of the Winner

When you buy a new thermos how do you know you are picking a good one? What criteria do you use? Size? Color? Brand reputation? If you are like me – probably some of all three.

But a good thermos (and I’m using lowercase ‘t’ here to mean consumer insulated liquid carrier) at it’s core should do exactly what its industrial cousin the dewar flask does – keep hot things hot, cold things cold, and keep you (the handler) from experiencing much of either when carrying it around.


Over the last couple years, through a combination of conference giveaways and borderline hoarding behavior, I’ve acquired a bunch of “high end” insulated drink carriers and used them with varying degrees of satisfaction. But earlier this month I laid out some cash too purchase what purported to be a VERY nice thermos-type drink carrier that I hoped might make my daily subway commute just a little more palatable.

What I realized when making my new thermos selection (yes, I’ve been humming that part from ‘The Jerk’ this whole time too) was that I had a woeful lack of actual data with which to inform my purpose. Basic things like brand, color and internal volume are available for pretty much any of these, but a thermos has actual performance metrics you can know and quantify.

Shopping for a thermos this way is as ludicrous as buying a car knowing only that it’s some shade of red, seats about 5 and has some number of horsepower.”

Don’t even get me started on the uselessness of most product reviews. Half of all product reviewers don’t seem to like anything at all, a quarter aren’t smart enough to know that of COURSE this thing wasn’t microwave safe, and the rest might provide some valuable insight if only I could isolate them from the others.

Most of these reviews and stats deal with qualitative data anyway – what I want to know is how well this thing will keep my drink hot.

So I designed an experiment to test the only real quantitative measure of insulated drink containers—heat loss.

Below I’ve detailed the experiment, the results, and provided a qualitative+quantitative review of each of the containers so that others looking for a great product wont have to stab in the dark and end up with a cupboard full of thermos like I did.


I tested the following drink containers:


  1. Standard Conference Giveaway Mug
  2. Sigg Wide Mouth Water Bottle (1L, uninsulated)
  3. Bubba Brands HERO Bottle(20oz)
  4. Thermos Vacuum Insulated Stainless Steel Hydration Bottle(18oz)
  5. Zojirushi SM-KHE36NL Stainless Steel Mug(16oz)


I attempted to create a hybrid approach to this test, compromising on a few points of control and accuracy in an attempt to keep the outcomes relevant to the real-world question I wanted to answer—which thermos is best? I’ll address the possible shortcomings of my methodology later, but I believe the test is more than sufficient to understand the thermal performance of each of the containers.

Test Steps:

  1. Set test-area temperature to 70 defrees (F)
  2. Heated water to 200 degrees (F)
  3. Filled each of the containers to its capacity with that water.
  4. Verified starting temp. of each container with an immersion-type digital thermometer.
  5. Recorded water temperatures over time with same immersion-type thermometer – opening the containers only to immerse the thermometer.
  6. Sampled at times that were relevant to the use-case (commuting and working in the office with the beverage)
  7. Logged those temperatures and their corresponding times to produce heat-loss curves for test group.


Conclusions: The best, the bargain, and the rest.

The Best: Zojirushi

The Zojirushi is unsurprisingly the winner in both this quantitative test and my qualitative observations. Even after 7 hours, the water inside was an impressive 142 degrees (F) – a full 30 degrees hotter than the next closest container. That’s the difference between a pleasant 70 degree day and a sweltering 100 degree day.

The Zojirushi was also my favorite to live with as a commuting mug. It is smaller and more slender than the others, the outside does not perceptibly change temperature when full of hot liquids (not surprising), and the 2-piece top is absolutely the best design I’ve seen.

My coffee stays HOT literally all day, the slender form factor means it can fit in a pocket or in my work bag without causing either to look like a snake that’s eaten an antelope.

It’s easy to clean, has not leaked a drop, and every detail reeks of the many hours it undoubtedly spent in some Japanese mechanical engineer’s CAD software.

The lock is secure and easily operated by one hand. When the top is opened (also a one-hand operation), someone has chosen the ‘correct’ spring tension for the lid. The lid opens and comes to rest in a set of grooves that positively hold it in the open position for drinking. Even more gratifying, when the lid does open, the fact that it is held in he open position is conveyed to the user with an inaudible, but tactilely perceptible little ‘pop’.

At $32, it is the most expensive container I tested (about twice the price of the closest performer), but like any good nerd, I think it’s worth the cost.

The Bargain: Thermos Vacuum Insulated Stainless Steel Hydration Bottle

If the Zojirushi is the BMW M3, the Thermos is a Ford Mustang GT.

At about $17, the Thermos did a yeoman’s work at keeping the water hot. Even after 7 hours, I would have described the liquid as “hot,” if not pleasantly so (and frankly after 7 hours it’s time to get fresh coffee anyway). The Thermos is relatively slender, holds a lot of liquid for its external volume, fits nicely in the hand, and has a locking top that is reasonably secure. I’ve owned 4 of these in the last 5 years and swore by them until I bought the Zojirushi.

There are, however, some major caveats to the Thermos’ performance. First and most notably; it is against the manufacturer’s recommendations to put hot liquids of any kind in this bottle (?!?!). Yes, thats right, this is called a “hydration” bottle because they want you to use it only for cold things. Why?

If I were to speculate, it would be because when filled with hot liquid and closed, the bottle becomes a pretty effective pressure vessel. If you snap the top closed before the air above the coffee (but below the mouthpiece) has had a chance to warm up, the next time you press the open button, the top springs back violently releasing a hot coffee-mist. I have never been burned, and the volume isn’t really enough to ruin anything, but it is annoying.

I get the feeling the plan was to make this a nice hot AND colt beverage thermos, but things went awry with the design of the top. Instead of redesigning the top, the team at Thermos just asked the marketing department to invent a term for “insulating beverage holder, but just for cold stuff” and called it a day.

The other shortcomings are with the gasket design and colorizing system.

The reason I’ve had 4 of these in the last 5 years is that I’ve repeatedly lost the single inner rubber gasket that keeps this sucker from leaking. After a few use cycles, it comes loose and will drop out during cleaning.

The top will screw closed and feel secure whether the gasket is in place or not. Without the gasket, the pressure vessel problem I mentioned before turns into a pressurized leaking problem. Hot beverage forces its way out and down the sides of the container, especially when it’s full – like, say, right as you step out the door in the morning and slip it into your clean coat pocket.

Where the Zojirushi is painted and clear-coated, the Thermos is colored using (what I assume is a much less costly) shrink-wrapping process. Basically, they make a plain stainless steel Thermos, and then place it inside a tube of transparent-colored plastic which is then heat-shrunk and adhered to the steel. While the tint is nice, it chips like thin nail polish. You can see what reasonable use has done to mine in the photo above. I honestly couldn’t care less about this, as I enjoy the patina that love brings to objects, but it is worth noting.

Meh: Bubba Brands HERO Bottle

This is the 1995 Camaro of insulated bottles.

It’s as tall as the Thermos, but almost TWICE as fat in the middle. It doesn’t perform as well, costs a little more, and gets hot to the touch around it’s collar. It does hold an extra 2 ounces of liquid, but they’ll be cold by the time you get there. The marketing materials tout its ability to keep hot things hot for 8 hours. It doesn’t. It touts its “sweat-proof’ design – presumably to keep it from slipping while it’s owner devours a second rack of ribs.

The others: The Sigg and the Mug

It isn’t really fair to compare these with the others that employ a vacuum as an insulating technique, but I figure’d they’d make good controls. The mug is like any other you’ve gotten for free—no manufacturer’s mark, probably emblazoned with some corporate logo, and made of something I’d qualitatively describe as “ceramic,” though who really knows.

This Sigg is a high-end steel hydration bottle (as opposed to the aluminum variants available) and is covered with what seems to be a orange-peel type powder coat. A nice bottle I use camping, but pretty much the opposite of insulated.

Notes on Methodology aka Pipe down, this is a customer review, not a scientific journal.

I’m acutely aware of a few major problems with my methodology – though I stand by my overall result as accurate enough to be useful. If I had infinite time and budget, here are a few of the things I wish I could have controlled for.

  • Age of the containers – The Zojirushi was relatively new at the time of testing, whereas the Thermos had obviously seen some shit. It’s likely that the Thermos would have performed better had it been as new as the Zojirushi, though I expect the differences would still be obvious.
  • Opening the container to record temps – Yeah, I did this with a nice kitchen thermometer, so the containers had to be opened one at a time to take the temps. Slightly different times remaining open means noise int he data. Also, I had top open them in series, not parallel, so there was probably a minute l between taking the temp of the first container and the last.
  • Climate control in the environment – I used my apartment’s thermostat for that, which I’d guess regulates temp +- 3 degrees. All the containers shares the same area when testing, so the noise should be uniform, but this isn’t ideal.

I hope this helps others who are looking to pick out the correct thermos. If you’ve had a different experience, or have some data to add to my analysis, feel free to hit me up on twitter.