We recently got to do something we'd been waiting years to do: bake side by side with three different varieties of Kernza®. One of them carried a seed about a third larger than the grain we normally mill. Another came from a brand-new Minnesota line. We cleaned all three, de-hulled them, milled them, sifted them, and baked our way from a 10% bagel up to a 100% Kernza shortbread.
Here's the short, slightly anticlimactic version: in the kitchen, we couldn't tell them apart. And that turns out to be one of the more encouraging things we could have learned about where this grain is headed.
Three varieties, one process. Can you spot the variety in the cracker? Neither could we.
The short version
- Kernza is a young grain that's improving fast. Plant breeders are selecting it for bigger seeds, easier threshing, and higher yield, and the pace of that breeding is accelerating.
- We tested three varieties — the established MN-Clearwater, the newer MN-Itasca, and the larger-seeded TLI-801 from Kansas.
- None of them required a single change to our process. Same mill setting, same de-huller, same sifting — even for the 34%-larger seed.
- In every baked good, from 10% to 100% Kernza, they were indistinguishable. Small wobbles in color and flavor pointed to the oven, not the variety.
- The one big difference was in cleaning — and we can't yet say how much of that is the variety, how much is the harvest, and how much is just one rough lot. We need more data.
- The lab found real differences between the flours. None of them showed up on a plate.
First, what exactly is Kernza?
Kernza is the grain produced by a perennial relative of wheat called intermediate wheatgrass. Unlike wheat, which is planted and harvested every year, a Kernza stand goes in the ground once and comes back for several seasons, sending roots ten feet deep. Those roots are the whole point. They hold soil in place, pull water and nitrogen down out of the watershed, and store carbon in a way no annual crop can match. It's the first perennial grain to make it to commercial kitchens and store shelves — and it happens to taste like honey, nuts, and brown butter, which doesn't hurt.
Kernza is a trademarked name. The plant underneath it, Thinopyrum intermedium, was first singled out for development as a food grain back in the 1980s, when researchers at the Rodale Institute went looking through dozens of perennial grasses for one with a seed big enough, and flavor good enough, to be worth eating. Intermediate wheatgrass won.
That history matters, because it tells you something about the grain we bake with today: it is still very early in its life as a crop. Wheat has had roughly ten thousand years of farmers saving the best seed. Kernza has had a few decades. Everything that makes it a little awkward in the kitchen — the small seed, the big bran-to-flour ratio, the gluten that's more "extensible" than "strong" — is a feature of a grain that's still being domesticated in real time.
How a perennial grain gets better
Here's the part we find genuinely exciting, and the reason we wanted to write this post.
Kernza is bred by a handful of public programs working together — most prominently The Land Institute in Salina, Kansas, and the University of Minnesota, which is where our established variety, MN-Clearwater, was developed. (MN-Clearwater, released in 2019, was actually the world's first intermediate wheatgrass variety bred specifically for people to eat.) These programs are all selecting for the same short list of traits, and if you've ever baked with Kernza, you can probably guess what's on it:
- Bigger seed — more flour, less bran, per kernel.
- Free threshing — seed that lets go of its hull easily, so de-hulling is less of a fight.
- Shatter resistance — heads that hold onto their seed until harvest instead of dropping it in the field.
- Higher yield — more grain per acre, the thing that ultimately decides whether a farmer can afford to grow it.
What's changed recently is the speed. For most of Kernza's history, improving it meant the slow, old-fashioned method: grow a field, walk it, pick the best plants, cross them, repeat. A single cycle could take three to five years. Then breeders started using genomic selection — reading a plant's DNA to predict which seedlings carry the good traits, before they ever reach a field. That single shift cut the breeding cycle down to about a year, and the program has even managed two cycles in a single year. The lead scientist at The Land Institute, Lee DeHaan, has put a number on where this could go: by his own extrapolation, Kernza could reach the current yield of Kansas wheat within roughly another thirteen years of breeding.
Take that with the appropriate grain of salt — it's an estimate, not a promise, and Kernza's yield today still sits at only around 10–25% of annual wheat. But the direction is unmistakable. The grain is a moving target, and it's moving toward us.
TLI-801, de-hulled. Its kernels run about 34% heavier than our Clearwater baseline — a preview of where breeders are steering the whole crop.
Why we ran three varieties through our whole process
All of that breeding raises a very practical question for a business like ours, which mills and sells Kernza flour and mixes every day: when a new, improved variety shows up, do we have to change how we work?
If every new line meant adjusting mill gaps, re-tuning the de-huller, and re-testing every recipe, that would be a real drag on the whole supply chain. So when we had the chance, through a Forever Green Initiative project led by food scientist Dr. George Annor at the University of Minnesota, to get our hands on multiple varieties at once, we ran them through our entire process and watched for anything that broke.
The three:
- MN-Clearwater — our commercial baseline, the variety we know best.
- MN-Itasca — a newer Minnesota line, with a seed about the same size as Clearwater's.
- TLI-801 — a Land Institute line out of Kansas, and the interesting one: its seed runs about 34% heavier than Clearwater's. It's the closest thing we had to a glimpse of "bigger-seeded future Kernza."
One honest caveat before the results. We had exactly one lot of each variety — the only material that exists yet — which means "variety" is tangled up with the field it grew in, the year, and the place. We can't fully separate the grain from its circumstances. So read what follows as a real-world first look, not a controlled trial. With that said, the findings were clear enough to be useful.
The headline: we didn't change a thing
The mill setting that handles our Clearwater handled all three varieties — same gap, same flow — including TLI's larger seed. The de-huller settings didn't move. Sifting came out effectively identical across the board. Whatever a new variety throws at a processor, none of these three threw anything our existing equipment couldn't take in stride.
Then we baked. We climbed a ladder of increasing Kernza inclusion, on the theory that if a difference between varieties existed anywhere, it would show up as you leaned harder on the grain:
Bagels at 10% Kernza — the first rung on the ladder.
Bagels (10%) came out as a matched set. A sandwich loaf (15%) gave us three nearly identical crumbs.
Pancakes at 20%. Any browning differences tracked the griddle, not the variety.
Pancakes (20%) browned a touch differently from plate to plate — but in a way that tracked the griddle, not the variety. Sourdough boules (~30%) rose and scored alike.
Scones at 41% — still no variety showing its face.
Crackers (40%) snapped into the same thin, even layers. Scones (41%) had the same tender, slightly bready crumb.
At 100% Kernza, with nowhere to hide, the three varieties still baked the same.
And then the test we'd built the whole ladder to reach: 100% Kernza shortbread. No wheat to hide behind, no inclusion rate to soften the comparison — just each variety, alone, as a cookie. If the varieties were going to diverge, here is where it would finally happen.
It didn't.
We did notice small wobbles — a little more color here, a little more flavor there. But the tell is that they never pointed the same direction twice. The variety that tasted mildest in one bake tasted boldest in the next. When the same grain reverses itself between two breads, that's not the variety talking. That's the oven, the fermentation, the time of day, and our own palates. It's noise, and noise is exactly what you'd expect to find once the real signal is this small.
The one real surprise: cleaning
There was one place the varieties came apart dramatically — but it wasn't in the mill or the oven. It was at the very first step, cleaning the raw grain of chaff, weed seed, and other material before anything else can happen.
The share of clean grain we recovered from each incoming lot:
- TLI-801: 78.5%
- MN-Itasca: 38.7%
- MN-Clearwater: 24.9%
That's a huge spread, and it's tempting to read it as "TLI is simply a cleaner, better grain." We don't think that's the whole story, despite TLI-801 having a meaningful increase in free threshed kernels. Here's what we think is really going on.
Some of this is about the individual lots, not the varieties. Each sample was grown by a different farmer, in a different place, and harvested a different way — and those circumstances explain some of the gap. The TLI-801 lot simply arrived cleaner: it came off a field managed closely for grain and was harvested in a way that left little chaff and weed seed to remove before we even started, and its larger seed makes whatever does come in easier to separate during cleaning. Our Clearwater lot, by contrast, came in with more wheat and other field debris mixed in — a rougher-than-usual incoming lot, and not representative of how Clearwater normally cleans, which runs closer to 35–45%. Itasca's lower number traced largely to camelina seed in that particular lot — a field weed, or a bit of carryover in the combine, not a feature of the variety. We removed it with an extra cleaning pass.
None of this is a verdict on which variety — or which region — grows cleaner Kernza. These are three single lots, each shaped by how it happened to be grown and handled that season.
Cleaning the raw grain — the one step where the lots came apart.
So the cleaning gap is real, and the potential payoff is real — cleaner incoming grain means less labor, more throughput, and lower cost per pound. But we're missing the one number that would let us settle it: yield per acre. Without it, we genuinely can't separate how much of TLI's cleaning advantage is the variety, how much is the way that particular lot was grown and harvested, and how much is simply its larger seed — most likely a mix of all three. We're calling it an open, very promising question rather than a verdict.
What the lab saw that the kitchen didn't
We also sent flour to Great Plains Analytical Laboratory for the full rheological workup — the same battery of tests we ran in our original deep-dive into Kernza. And here the two new 2025 lines, TLI and Itasca, did show real, measurable differences from each other.
Itasca turned out to be the stiffer, more mineral-rich flour, with starch that swells to roughly twice the viscosity of TLI's. TLI was the more balanced and extensible of the two, with more intact starch. These aren't rounding errors; they're genuine differences in personality between two varieties.
And not one of them surfaced in anything we baked.
That's the whole story in a sentence: the lab tells us these varieties are not identical, and the kitchen tells us it doesn't matter — at least not at any inclusion rate a real baker actually uses. The instruments resolve differences our process simply cannot feel. Both things are true at once, and it's worth holding them together: a variety can be measurably distinct on a bench and completely interchangeable on a plate.
What this means for the future of Kernza
Step back, and a picture comes into focus that we find genuinely hopeful — with two honest caveats.
The hopeful part: the traits breeders are chasing are exactly the traits that make Kernza easier and cheaper to turn into food. Bigger seed means more flour and less fighting with bran. Free-threshing seed means easier de-hulling. Cleaner, more uniform grain — whether from genetics or better harvest practices — means less waste at the very first step, which is where TLI quietly ran away from the pack. Every one of these improvements pushes the cost of Kernza down. And cost is really the bottleneck right now. Kernza is still expensive to grow and process, which keeps its price well above the grains it competes with — and that price, more than any shortage of grain, is what holds the market back. Cheaper grain means lower prices, and lower prices are what will bring more bakers, businesses, and home cooks to the table. The breeding pipeline is a big part of how we eventually get there.
The first caveat is something we honestly don't know yet: what a bigger seed does to nutrition. Much of what makes Kernza so nourishing — its fiber, its minerals, its protein — rides in the bran, and Kernza carries so much bran precisely because its seed is small. Breed the seed bigger and you add more of the starchy endosperm in the middle, which is part of what makes a grain bake more like wheat in the first place. That might turn out to be a wash, or it might dilute some of the bran-borne nutrition that sets Kernza apart, gram for gram. We don't yet know how it will present on a nutrition label — and it's something we'll be watching closely as bigger-seeded lines arrive.
The second caveat is more familiar: a fast-improving grain is a moving grain. New varieties will keep arriving, each a little different from the last. For a processor, that means we can't ever assume a variety is the variety forever — we'll need to keep checking, keep milling test batches, keep tasting. That's real work, and it never quite ends.
But our test suggests the work is far lighter than we feared. The single most reassuring result we got is that a seed 34% larger than our baseline dropped into our existing process without a single adjustment. If that's what the near future of Kernza looks like — bigger, cleaner, higher-yielding, but still slotting into the equipment and recipes we already run — then the supply chain can absorb a generation of improvement without flinching. Challenging, yes. But challenging in the best possible way: the grain keeps getting better, and we mostly just get to keep baking.
The bottom line
We set out to answer a narrow, practical question — does a new Kernza variety force a food processor to change anything? — and walked away with a broader piece of good news about the grain's whole trajectory.
For now, the takeaway needs no caveats to stand: we can bake with these new varieties today, exactly as we bake now, and our customers will not be able to tell the difference. The lab can see the future arriving. Your kitchen won't have to.
Kernza is still young, still improving, and still — to our continued delight — one of the most genuinely hopeful things you can put in an oven. Bake more of it.