How Dulce de Leche Works

Reader Kitty writes:

So the whole putting a can of condensed milk in a pan of boiling water thing makes caramel because of the pressure cooking and thus shortens the time? Or perhaps they are a bit more base as well… hmmm.

Hey Kitty! It depends on the technique you use to make dulce de leche. If you do it the old-fashioned way by simply simmering milk and sugar for an hour or so, you get Maillard-type browning initially, and then caramelization. If you boil a can of sweetened condensed milk in a saucepan, you just get caramelization.

I know what you’re thinking: Hang on a second Joe, doesn’t it take a lot more heat to create caramelization…like 300 degrees Fahrenheit or more? Boiling water can only reach 212 or so before it turns to steam. What gives?

Well I’ll tell you. It’s true that sugar can’t caramelize at the temperature of boiling water. However there’s an area of a pan of boiling water where the temperature greatly exceeds 212 degrees Fahrenheit: the bottom. It’s sitting right on top of the heat source. When you insert a can of sweetened condensed milk into that pan, the bottom of the can is in direct contact with the bottom of the pan, which is sitting right on top of the heat source. Down there, temperatures get up over 300, and that causes the sugars in the lower part of the can of sweetened condensed milk to caramelize.

So really, it’s not the boiling water that’s creating the caramel. In fact the boiling water is actually cooling the can so it doesn’t blow up. The method works most of the time…though not all of the time. Sometimes people boil the can for so long that the water level drops and exposes the can. Other times you get steam buildup and an explosion all the same. Either way the result is a shower of boiling water and caramel for whomever is standing in the kitchen at the time. And zis, she iz no good.

Some folks get around these dangers by poking holes in the top of the can to relieve the pressure. They then fill the pot only to the lip of the can. That’s probably a better solution, but still not something I’d ever recommend. The best method I know of for converting sweetened condensed milk into caramel is to open the can and pour it out into a baking dish, then place the dish in a water bath and bake it in a 400+ oven. This way you get good caramelization of the surface while the pan stays cool enough so the caramel won’t burn. Having done it, I can say that it works.

The problem for me is that no matter which method you use for caramelizing sweetened condensed milk, your raw material is still sweetened condensed milk, and that produces a vastly inferior dulce de leche, at least in my view. The simmering method allows you to mix your milks, get creative with your sugars, add spices, whatever you like. And it doesn’t take a whole lot more time.

28 thoughts on “How Dulce de Leche Works”

  1. I’ve successfully cooked cans of condensed milk in the crock pot to make dulche de leche. 8 hours makes a pretty dark version, if I ever did it again, I’d go 4-6 instead.

    Me not being fond of scraping caramel off my ceiling, I put my crockpot outside. Just in case.

    1. Probably a good solution, GL. But try my recipe, I think it’ll make a convert out of you!

      – Joe

    2. I have made dulce de leche in a slow cooker, but not with sweetened condensed milk. I used milk, sugar and soda and cooked as I would on the stovetop, but for a lot longer, stirring when I remembered.

      1. Hey Mary, that’s a nice method. I’m going to try that as well.


        – Joe

  2. Eh I’ve never actually done it, I’m just curious… obviously the real stuff is better than canned, isn’t it always? But I can’t say I’m not interested in doing it just for the nuttiness OF doing it. and then giving a can or two away 😉

  3. I have made Joe’s dulce de leche recipe, which he links to in the above post, and it’s one of my favorite recipes on the site, both easy and delicious. Follow his advice; you WILL be a convert.

    1. Thanks for the boost here, RL! As you can see I’ve got quite a few readers here risking life and limb for an inferior product.

      – Joe

  4. I always thought the can was being a pressure cooker. The lady with the crock pot doesn’t have a hot bit at the bottom.

    1. Interesting point. My guess is that the bottom of the crock pot is getting hot enough. I’m not saying I know for certain that pressure doesn’t aid caramelization of sugar, but if it does, why don’t other foods that are pressure canned show evidence of it?

      I’ll see what I can dig up on this.

      – Joe

    2. I’ve been thinking more about this, and it seems to me that sweetened condensed milk is basically dairy fat (about 25%), milk potein (about 10%), and sugar syrup. I’m not too sure to what extent that syrup has been reduced, but considering that the product is “condensed” my guess is that it’s been reduced a fair amount. Which means the boiling point of that mixture must be well above 212 F, which is the temperature of the surrounding water. Given that you need steam to create pressure inside the can, my guess is that there isn’t very much, at least when the can is covered its water. However when the water level drops and the can is exposed, there’s almost no limit to how hot the insides of the can can get. The results is then steam and an explosion of some very, very hot liquid. Or at least that’s how it seems to me. What do you think?

      – Joe

      1. Don’t know. Could open a tin of condensed milk and take its temperature while bringing it to the boil I suppose. See what the boiling point is empirically.

  5. How long can we keep a can of DdL? One which I have boiled in a saucepan for 4 hours?

    1. I would think almost indefinitely. As long as sweetened condensed milk at least. Not that I’m recommending it, of course.

      – Joe

    1. Fabulous, Eva! It looks fantastic!

      I’m very glad to have someone try it, filo and all. Wish I could have been there for a bite! Cheers,

      – Joe

  6. A pressure cooker works really well for canned sweetened condensed milk (and it’s safe- pressure outside the can keeps pressure inside the can from exploding). I have an electric Cuisinart pressure cooker. I usually put an unopened can in with a 1/2 cup of water for about 30-45 minutes. An hour will produce a caramel with the consistency of a very thick custard and is very dark; tastes good but you can’t pour it.
    *Let the can cool down a little before opening. It will squirt HOT caramel out if opened too soon.

    I wonder if putting your recipe in a mason jar inside the pressure cooker would work?

    1. Oh good Lord, you trying to give me a heart attack, John? 😉

      NO, definitely don’t try putting my mixture into a mason jar. In an hour you’ll have better dulce de leche than you’ve ever had in your life — and no scars to show for it!

      – Joe

  7. I’ll hve to try your oven method. I boil the can. The one time I opened the can and tried the stove-top method I was very disappointed. Boilinghte can has always worked for me with no ill effects… but the warning label scares me every time.

    1. See what you think of the oven method…but regarding the stove top, did you try the recipe here on the site?

      – Joe

      1. I’m embarassed to say that I did not. I followed someone else’s recipe. No wonder it didn’t work! I’ll try yours ASAP. Just reading the ingredient list makes me certain that yours will taste much better than boiling the can.

        1. Give it a try and let me know, Brian. I’ve had a lot of success with this, but I want a full report! 😉

          – J

  8. So Joe, can you enlighten me where in a pan are you supposed to take a thermometer reading? I was making a lemon tart (Heston Blumenthal’s recipe sorry) at the weekend and it says to take the custard to 62C, I was using my thermometer but depending on where I put it I got different (like much different) readings. I was bringing this custard up over a pan, and I must have stood there for an hour stirring stirring stirring until I finally thought “oh blow it” and bunged it all in the tart case and put it into the oven…it came out perfect by the way, but what a faff!

  9. Thanks so much. I have been looking for a recipe NOT in a can for a bit now. Canned food is so bad for the body. Thank you thank you!

  10. This reaction appears to take place at room temperature given a sufficient amount of time. I just opened a can of sweetened condensed milk that was “best if used by” dated March 2007. The entire contents is perfect dulce de leche, at least as perfect as you can get starting with condensed milk. The can was stored on a garage shelf with seasonal cycling between single digits and 90’s. It just comes down to reaction kinetics! Hate to wait 10 years every time I wanted to make some but it was a nice surprise to find that I didn’t have to lift a finger to enjoy the end result for once.

    1. Hey Greg!

      Fascinating. The last few years I’ve read on and off about slow-motion low-temperature Maillard reactions that happen at those temperatures. Really beats the conventional food world wisdom that it takes a lot of heat to create browning. Instant browning maybe…this is really fascinating science.

      Cheers and thanks for the note!

      – Joe

Leave a Reply to Brian Shaw Cancel reply

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