Whatâs the best way to cook a sweet potato? I donât mean those boiled-to-death, mashed-with-maple syrup, drowning-in-marshmallows sweet potatoes (no shade at sweet potato casseroles). How do you cook whole sweet potatoes straight up, with little adornment?
Some people boil their sweet potatoes. Some bake them. Some even use the microwave. But in the interest of maximizing flavor, texture, and convenience, which cooking method comes out on top?
In the past, Kenji showed us that both low-temperature baking and sous vide cooking were reliable methods to maximize sweetness in sweet potatoes. And this year, Chef Lucas Sin of Junzi Kitchen and Nice Day Restaurant went viral for his take on a frozen roasted sweet potato, inspired by his experiences growing up in Hong Kong.
So which is it? Boiled, baked, sous vide, frozen and baked, or something else? Letâs find out.
A Spud By Any Other Name
Sweet potatoes arenât really potatoes at all (though they’re distantly related). Theyâre part of the morning glory family (Convolvulaceae), of which the most familiar varieties may be those spindly, vining plants that explode with multi-colored, spiral-shaped buds from summer to fall. In fact, in many parts of the world, sweet potato vines are grown solely for their ornamental flowers, which look similar to those garden-variety morning glories that might adorn your backyard.
According to some sources, sweet potatoes rank among the most important and productive food crops in the world, after rice, wheat, maize, and sorghum. The part we eatâthe large root or tuberâis sweet, starchy, and prized for its nutritional value. Most people cook their sweet potatoes, which further increases nutritional value and digestibility. They can be boiled, steamed, baked, fried, or otherwise subjected to any kind of heat.
Letâs go over a few general points about sweet potatoes before getting too deep. To paraphrase J. Kenji Lopez-Alt, SEâs original spud god: Yams are not sweet potatoes. Yams are starchy tubers that belong to a completely different genus of plants (Dioscorea) and can range in size from as small as a baseball to a hulking mass upwards of 100 pounds. They also taste a little different than sweet potatoesâtheyâre generally drier, starchier, and not so sweet.
To quote Kenji once again, hereâs a rough breakdown of the types of sweet potatoes you might find at the store:
Dry sweet potatoes, like the white-fleshed Jersey sweet potato or Okinawan purple potatoes, are starchier and less sweet than moist sweet potatoes. They turn fluffy when cooked, and in many recipes can act as a good substitute for normal potatoes, albeit with a unique flavor all their own. Moist sweet potatoes, like the Garnet or Ruby âyam,” or the Beauregard sweet potato, are the more widely available variety in the US. They have a higher water content and sugar content than dry sweet potatoes, and they cook up creamy and rich, rather than fluffy.
And if you want the full primer on sweet potatoes, yams, and everything in between, Elazar Sontag wrote a great field guide.
Given the vast catalogue of sweet potato varieties, itâs important to know whether the âbestâ cooking method applies to many of these varieties, not just one. That meant running multiple sweet potato types through our tests to make sure our recommendation was solid across the board.
What Is a “Perfectly Cooked Sweet Potato”?
Regardless of method, what are the qualities we look for in a well-cooked sweet potato? What are the desired flavors and textures, and what are the dealbreakers? Many of these parameters are subjective, of course, but itâs still useful to think in these terms as we move along.
- Sweetness: A cooked sweet potato should taste relatively sweet (itâs in the name, after all). Kenji went to great lengths to explain sweetness in sweet potatoes: they contain an enzyme (amylase) that converts the tuberâs complex starches into maltose, a basic sugar molecule;, and prolonged cooking encourages this enzymatic action. He showed that the more starch you convert to maltose, the sweeter the sweet potato tastes. But more on that later.
- Aroma: There are a number of volatile compounds that contribute to the nutty, earthy, caramel aroma we associate with cooked sweet potatoes. For instance, this study shows that there are around 60 volatile compounds present in cooked Jewel sweet potatoes, of which three are the most apparent: phenylacetaldehyde (perfume), maltol (caramel), and methyl geranate (sweet candy).
- Do we necessarily care about all those compounds? Not exactly. We probably just care that our cooked sweet potatoes are intensely aromaticâanything but plain or bland. And as we will see, the right cooking method can have a profound impact on that aroma.
- Texture: Depending on the variety, cooked sweet potato can have textures ranging from fluffy and crumbly to wet and mushy. Because of that variability, itâs difficult to make blanket statements about desired textures. It might be more useful to identify what we don’t want in a cooked sweet potato: Gumminess; dry, leathery spots; crunchiness; or uncooked sections. At the very least, it might be better to hedge toward a softer texture, regardless of variety.
Maltose: The Secret to Sweet Potato Sweetness
One of the keys to a truly sweet sweet potato is to maximize the conversion of its naturally occurring starches (which we don’t perceive as particularly sweet) into their candy-sweet, sugar building blocks. In this case, that sugar is maltose.
The main driver of this starch-sugar conversion is an enzyme called amylase. But there’s a catch: amylase only does its best work in a specific temperature rangeâbetween 135Â°F and 170Â°F (57Â°C and 77Â°C). Holding a sweet potato in this temperature zone for a longer period of time yields a sweeter tasting finished product, and itâs the key to Kenjiâs recipes for Mashed Sweet Potatoes and Roasted Sweet Potatoes.
Still, thatâs not the full story. Sure, more maltose means more sweetness. But more maltose also means more aroma, and therefore more sweet potato flavor. Thatâs because maltose is a primary precursor for several of those crucial volatile compounds we find in cooked sweet potatoes. In other words, maximizing maltose maximizes both sweetness and overall flavor. So if thatâs the goal, how do we approach cooking?
I tested six cooking methods and compared results. I cooked two pounds each of whole, scrubbed sweet potatoes (Garnet and Hannah varieties) and gave them the following treatments:
- Boiling: I kept the sweet potatoes submerged in water at a low simmer until they were fork-tender at the core.
- Baking/Roasting: Following Kenjiâs method for baking sweet potatoes before mashing them, I wrapped sweet potatoes in aluminum foil and roasted them at 300Â°F on the middle rack until they were fork-tender at the core.
- Steaming: I steamed whole sweet potatoes in a bamboo steamer until they were fork-tender at the core.
- Microwave: I microwaved whole sweet potatoes in a covered bowl until they were fork-tender at the core.
- Sous Vide: I vacuum-sealed whole sweet potatoes and cooked them in a 150Â°F water bath for three hours.
- Lucas Sinâs Freeze-Roast Method: Sin calls for freezing whole, scrubbed sweet potatoes solid, then roasting them uncovered at 450Â°F until they bubble at the surface with burnt sugar and they’re fork-tender at the core. This method is based on his memories of passing street vendors as they repeatedly shoveled hot coals over smoky, fragrant sweet potatoes. As Sin explains, in certain parts of China, itâs cold enough to freeze the sweet potatoes just by leaving them outside, which results in a fluffy texture; the hot coals produce a smoky, mildly charred exterior and caramelize the sugars inside without burning them.Â
With the exception of the sous vide treatment, all of these methods produced comparable textures: soft, tender, and plump for the moist Garnets; soft, tender, and fluffy for the dry Hannah variety. The freeze-roast method seemed to yield a slightly softer texture overall, and it was apparent that the frozen sweet potatoes lost more water during roasting compared to the other methods, as each frozen sweet potato produced a syrupy liquid that pooled out onto the baking sheet as it cooked.
The sous-vide method yielded very firm sweet potatoes, and that was to be expected: pectin doesnât really break down until temperatures above 180Â°F (82Â°C).
Flavor and Aroma
The biggest differences were in this category. If I had to rank the methods in order of flavor, it might look something like this:
- Sous vide
The top two methods yielded sweet potatoes with plenty of sweetness and classic, earthy-sweet, caramel-like aromas. The sous vide method resulted in similar sweetness, but I found it produced sweet potatoes that were significantly less aromatic.
In comparison, steaming,Â boiling, and microwaving all produced relatively bland sweet potatoes with little to no aroma. Steaming was marginally better, and I could taste hints of sweetness in the samples. Microwaving was the worst method by far: while the texture was passable, these samples tasted like watered-down cardboard, a mere suggestion of sweet potato rather than the real deal.
So how do we explain the results?
Volatility Is Key
As it turns out, baking or roasting really might be the best method overall. This study compared baking, boiling, and microwaving sweet potatoes, and showed quantitatively that baking produced significantly higher numbers and concentrations of desirable volatile compounds and Maillard products than any other method. The authors explain that boiling and microwaving appear to deactivate those critical amylase enzymes in sweet potatoes long before they have a chance to convert starches to maltose (which is a precursor to the volatile compounds weâre after). The result is blander-tasting sweet potato. Water is a really good thermal conductor, so boiling facilitates rapid heat transfer, which quickly raises the internal temperature of the sweet potatoâhigh enough to deactivate amylase as well as other crucial enzymes. Similarly, microwaving is a fast cooking process in which heat is generated throughout the sweet potato at approximately the same rate.
Baking or roasting is an imperfect cooking process in which heat moves from the exterior to the interior. The combination of imperfect heat conduction and low humidity means that the sweet potato spends more time in that sweet spot temperature for amylase to work its magic, converting starches to maltose. Over time, that maltose undergoes Maillard reactions because of the high heat of the oven, and eventually caramelization occurs. Thatâs why you get a fuller spectrum of aroma and flavor through baking or roasting.
Putting Things On Ice
Ok, weâve established that the oven is, in fact the way to go. But given my results, I still found Lucas Sin’s method of freezing and then roasting to be slightly superior to roasting alone. So what gives?
Letâs think about the freezing process for a second. As with most vegetables, sweet potatoes have appreciable water content. When we freeze them, water molecules expand; when those molecules thaw and revert to a liquid state, the cell walls of the vegetable tend to rupture, and water seeps out to the exterior, turning the interior texture soft. This process is the crux of cryo-blanching vegetables, which Kenji talks about here.
We see this process in action with Lucas Sinâs method. Water quickly pools out of the sweet potatoes if baked for long enough, along with maltose. (According to Sin, driving out water is a key concept in Chinese cooking; it is thought to concentrate flavor.) Over time, this sugary water caramelizes and even burns. So the interior, according to Sin, is less watery, and more concentrated in flavor.
The second and arguably more important explanation is this: Freezing lowers the starting temperature of the sweet potatoâso much so that it spends even more time in that ideal lower temperature range for amylase activity as it slowly heats in the oven. More amylase activity means more maltose, which means more potential for Maillard reactions and production of volatile compounds, i.e. more flavor. (If I could run a supplementary test to confirm this idea, I would stick a temperature probe into a frozen sweet potato as it baked, and record the internal temperature over time. My guess is that the temperature gradient would be lower than that of a non-frozen sweet potato, and we would expect a slightly longer period of time spent between 135Â°F and 170Â°F.)
Itâs the same reasoning behind starting the sweet potatoes in a cold ovenâa technique which Kenji employs in order to maximize maltose conversionâbut taken to an extreme.
Playing With The Concept
Now that we understand why Lucas Sin’s freeze-roast method works so well, it opens up a world of possibility. There’s no one ideal way to employ the technique, but there are levers to push and pull depending on how much time you have. Have some sweet potatoes on hand and no immediate need to cook them? Throw them in the freezer until frozen solid. In a rush? Even a shorter stint in the freezer to just partially freeze the flesh will help it spend more time in that ideal lower-temp zone during roasting, even if it isn’t quite as ideal. Here are some other ways to play with the method:
Start It Cold, Even in the Oven
If you have less time, you may want to preheat your oven while the sweet potatoes ice over in the freezer; that way you can just chuck the semi-frozen tubers into the hot oven and have them ready faster. But if you can spare an extra hour or so, why not start the frozen potatoes in a cold oven, as Kenji does here? The gradual heating gives even more time for enzymes to do their job, resulting in a sweeter, more aromatic sweet potato.
Start Low, Then Crank The Temp
Building on the idea of starting in a cold oven, we can manipulate the oven temperature further. Instead of starting cold and then letting the oven quickly climb to the final high roasting temperature, we can set an initial lower temp of about 300Â°F, followed by a finishing roast at 450Â°F. The lower roasting temperature at the start means more time spent in the enzymatic sweetening phase, while the final blast of high heat drives off any excess moisture, and produces those charred, smoky, caramel-like flavors that define Sinâs method.
Wrap It Up, Then Take It Off
Wrapping sweet potatoes in foil helps in a few ways. It prevents the exterior of the tuber from drying out and overheating too quickly, which would minimize enzymatic activity; it also results in a more evenly cooked texture. Finally, you can incorporate aromatics in the foil package to jazz things up.
Wrapping in foil is great for the initial cooking and flavor development. But Lucas Sinâs recipe is all about generating that charred, caramel-like flavor. By unwrapping the sweet potatoes and continuing to bake them, all that maltose is free to undergo further Maillard reactions and caramelization.
Place sweet potatoes in freezer until frozen solid, at least 6 hours and up to 3 months
Line a rimmed baking sheet with aluminum foil. Wrap each frozen sweet potato in aluminum foil, adding aromatics if desired, and seal tightly so that there are no loose seams. Place wrapped sweet potatoes on prepared baking sheet and place in oven. Set oven to 300Â°F (150Â°C). Roast until a thin skewer inserted into potatoes meets no resistance, about 2 hours.
Remove baking sheet from oven and carefully unwrap potatoes; discard aromatics, if using. Increase oven temperature to 450Â°F (232Â°C) and return unwrapped sweet potatoes to oven. Roast until sweet potatoes are fragrant and sugars begin to caramelize, about 40 minutes. Let roasted sweet potatoes stand until cool enough to handle, then serve as desired.
This recipe works equally well with dry or moist sweet potato varieties. Dry sweet potatoes such as Okinawan purple or Hannah varieties will be fluffy and less sweet. Moist sweet potatoes such as Jewel, Garnet, or Beauregard varieties will taste juicier and sweeter.
Make-Ahead and Storage
Raw sweet potatoes can be kept frozen, unwrapped, for up to 3 months. Cooked and cooled sweet potatoes can be stored in the refrigerator for up to 1 week. You can also freeze the cooked potatoes, wrapped in plastic and foil, for up to 1 month.