On 21 September 2026 I read four English guides to this question, all of them written by shops that sell Japanese knives: Knives and Stones, Hasu-Seizo, Oishya and Blade and Board in New Zealand. Each one is linked below, at the point where I quote it.
On your first stone they agree with each other, and they agree with Japan. Buy a 1000 grit stone. Blade and Board puts it in six words: if you only buy one stone, buy a 1000 grit.
That part of the question is settled. This article is about what you buy second, because that is where the English answer and the Japanese answer come apart, and the gap costs about sixty dollars.
The English answer is a ladder. Go up: 3000, then 6000, then 8000, a polished edge. The Japanese trade has a different second step, and it is not a grit at all.
I read three Japanese sources this week. A knife shop that sharpens for the chefs of the old Tsukiji market. A knife maker in Niigata. A whetstone manufacturer. All three, independently, name the same thing as the most important, and none of them names a grit number.
This is the tenth piece in a series about the distance between what Japanese knives are called abroad and what they are in Japan. The one before it took apart single bevels.
First, the thing nobody tells you when you buy the knife
Your new Japanese knife is not finished. The shop I have a partnership with says so in its own help pages:
Normally, a obtuse angle finishing at the blade edge by a machine in manufacturers, called standard blade finishing, is applied to most of the knives that are stocked and displayed in the store.
And then:
Professional chefs often apply a hand-finishing to the blades by themselves just after they buy the knives.
Read that twice. The edge on a shelf knife is a machine edge, ground obtuse on purpose so that it survives whatever the buyer does with it. A professional treats the purchase as half the job and puts the knife on a stone before the first service, not after the edge dies.
Tojiro, the maker in Tsubame-Sanjo, treats this as the central problem of selling Japanese knives abroad:
藤次郎は日本の包丁を海外などで販売する上で「研ぎ直し」を一番重要なことと考えています。販売した先で刃のメンテナンスができなければ日本の包丁の良さは維持できないからです
For selling Japanese knives overseas, Tojiro considers re-sharpening to be the single most important thing. If the edge cannot be maintained where the knife is sold, what is good about a Japanese knife cannot be maintained either.
It backs that with money. Where a national distributor has no sharpening skill, Tojiro says it has them install a water-cooled grinder and trains them on it. So the stone is not an accessory to the knife. It is the half of the product that gets shipped to you unassembled.
The first stone: everybody agrees, and the Japanese sources are blunter
Tojiro Knife Gallery, asked by a whetstone maker what customers should buy, answers in one line:
包丁研ぎは中砥石が一番重要ということもあり、クリーンセラの中砥石がよく購入されていますね
Because the medium stone is the most important thing in sharpening a knife, the Cleanthera medium stone is what sells.
The 中砥石, the medium stone, is the 1000 grit stone. Tojiro sells it in two sizes, 210 x 65 x 32mm and 210 x 70 x 60mm, both 1000 grit.
Tsukiji Aritsugu, a knife shop that has supplied the fish market trade for generations, gives the same number for ordinary customers, and then adds a stone that the English guides never discuss:
一般の方だと中砥石の剛研デラックスの#1000がよく出ますね。あと青砥石の評判がいいね
For general customers the Gouken Deluxe 1000 grit medium stone sells well. And the aoto has a good reputation.
青砥石は良く研げて特にツヤの出方がちょっと違う。粒度も#2000だけど仕上砥石としても十分使える砥石です
The aoto cuts well and the lustre it leaves is a little different. The grit is 2000, but it works perfectly well as a finishing stone too.
A Tsukiji knife shop says 2000 grit is enough finish for a kitchen knife.
Compare that with the English shops. Knives and Stones gives the standard beginner kit in one sentence: if you buy one stone, make it a 1000 and 6000 combination, the 1000 side sharpens and the 6000 side polishes. Hasu-Seizo draws its line lower than most, saying most Japanese knife users only need 1000 to 3000 grit for outstanding results. The Tsukiji retailer puts it lower still, at 2000 grit, and calls that a finishing stone.
What the Tsukiji shop asks back
Here is the sentence this whole article exists for. Asked for advice on sharpening, the president of Aritsugu says:
正しく研ぐことは大切なんだけど、一番大切なことは砥石の面を常に平らにすることですね
Sharpening correctly matters, but the most important thing is to keep the face of the stone flat at all times.
Then he explains how he knows when someone has not:
上手く研げないと持ってきた包丁を見ると凹んだ砥石を使っている事がわかります。凹んだ砥石だと鋭角な刃が丸く蛤刃になるから切り刃の幅が狭くなる
When someone brings in a knife they cannot sharpen well, I can see from the knife that they are using a dished stone. On a dished stone an acute edge turns round, into a convex hamaguri edge, so the width of the bevel narrows.
That is a diagnosis from the knife, not from the stone. The shop does not need to see your stone. The hollow in your stone is written on your blade, because a hollow stone grinds a curve into the edge and the flat bevel you were trying to make gets narrower every session. Your knife stops taking an edge, you conclude you are bad at sharpening, and you buy a finer stone to fix it. The finer stone has the same hollow.
His fix costs nothing and has no brand in it:
面直し砥石やコンクリートブロックでもいいので削って凹みを直すことが大切です
Use a flattening stone, or a concrete block would do, and grind the hollow out.
And the frequency is not what you would guess:
だから研ぎ終わったら都度、砥石の面を平らにする事が大切です
So it matters to flatten the face of the stone every single time you finish sharpening.
Not when it looks bad. Every session.
A second Japanese shop, same complaint, different city
Tojiro Knife Gallery, 300km away and in the business of making knives rather than selling other makers, reports the same failure in its customers:
砥石表面の歪みについて知らないで歪みを直してない方が多くいます。家にある砥石について聞くと「砥石は持ってる、家にあるんだよ。でもね、凹んでるんだよね。」て
A lot of people do not know that the surface of the stone distorts, and never correct it. When I ask about the stone they have at home, it is: I have got a stone, it is at home. But you know, it is dished.
The gallery keeps a ruined stone on the counter to show people. Two shops, in different trades, in different prefectures, volunteer the same diagnosis to the same question.
And the company that makes the stones
Shapton, a whetstone manufacturer in Tochigi, writes it into the sales copy for its flattening plate:
刃物を研ぐ事により砥石の変形が始まり、研ぐ回数を重ねる度に、刃物の平面までも損われていきます。大切な事は、砥石を平面に保つ事です
Sharpening a blade starts to deform the stone, and with every further sharpening the flat of the blade itself is damaged as well. The important thing is to keep the stone flat.
That is the mechanism stated by the party with the least interest in stating it. A stone maker would rather sell you the next grit.
You can also count its product line, which is harder to argue with than any sentence. Shapton sells four separate things whose only job is the condition of the stone face: a glass lapping plate, a double-sided diamond plate called Kubo, a grooved flattening block called Naoru that comes with correcting powder, and a small dressing stone for a clogged surface. Four products for the stone, against one ladder of grits for the knife.
What that does to the shopping list
Prices below are from Hocho-Knife on 21 September 2026, all in stock when I checked. I have an affiliate partnership with them, explained at the bottom.
The English ladder sends you here for your second purchase:
- Shapton Kuromaku 8000 grit, the melon-coloured one, $90.99 — a finishing stone, 210 x 70 x 15mm.
All three Japanese sources send you here instead:
- Naniwa flattening stone, $27.99 — 170 x 55 x 25mm, 250g. It sharpens nothing. It makes your 1000 grit work as designed, every session, for as long as you own it.
Sixty-three dollars apart, and the cheap one is the one that fixes the problem you actually have. The finishing stone is not a bad product. It is the third purchase being sold as the second.
How much stone are you actually buying
Once you accept that the face gets ground back every session, thickness stops being a specification and becomes the quantity of sharpening you paid for. The numbers are rarely put side by side, so here they are, same shop, same day:
- King 1000 grit Middle, $45.99 — 207 x 66 x 34mm, 500g.
- King 1000 grit Large, $62.99 — 209 x 70 x 60mm, 940g.
- Naniwa Super Ceramic S-1000 with stand, $75.99 — 210 x 70 x 20mm, 800g.
- Shapton Kuromaku 1000 grit, $53.99 — 210 x 70 x 15mm.
- Tojiro combination 1000 and 3000 with stand, $44.99 — 170 x 55 x 24mm total, two grits sharing it.
The combination stone is the interesting one. It looks like two stones for the price of one, and at $44.99 it is the cheapest thing on the list. It is also 24mm thick in total, which is less than the 34mm of the single King 1000 grit that costs a dollar more. Split that between two faces and each grit gets about half. Flatten both faces for a few years and there is not much left.
Aritsugu makes the same point from the other end, about the aoto he likes:
それに青砥石は厚みがあるから一般家庭だと一生ものでしょ
And the aoto is thick, so for an ordinary household it is a lifetime stone.
The aoto he means is 55mm thick. He adds a safety point that no online guide I read mentions:
薄くて軽い砥石は研ぐときに動きやすくて危ないからね。初めのうちはどっしりとした安定した砥石の方が安全
A thin, light stone moves around while you sharpen, which is dangerous. Early on, a heavy stable stone is safer.
One honest caveat, because thickness alone is not lifespan. A stone that cuts fast usually wears fast. Shapton claims the opposite trade for its ceramics, that optimising the binder gives ヘリが少ない, low wear, and its thin 15mm stone is built on that claim. Its top line is thinner still: the Rock Star is 10mm, of which the abrasive layer is 5mm, and the maker prints that number itself. So compare thickness within a family, not across one. What you should never do is read 24mm on a two-sided stone as if it were 24mm of sharpening.
The grit number on the box is not a unit
There is no shared standard behind these numbers that lets you compare a 1000 grit stone from one country with a 1000 grit stone from another. What you can do is use a maker that publishes the particle size. Shapton prints the micron figure for every step of its Rock Star line, and the spacing is worth seeing:
- 320 grit is 40 micron.
- 500 grit is 25 micron.
- 1000 grit is 11 micron.
- 2000 grit is 6 micron.
- 3000 grit is 4 micron.
- 8000 grit is 1 micron.
From 320 to 1000 the particle size falls by 29 micron. From 3000 to 8000 it falls by 3. The ladder is not evenly spaced, and almost all of the real work happens before 1000 grit. That is the same conclusion the Tsukiji shop reached by hand, and it is why a medium stone plus a flat face beats a shelf of finishing stones.
The angle question, which none of the four guides raised
I went looking for the 15-degrees-per-side advice that circulates online and found something odd: not one of the four English grit guides mentions an angle at all. The shop that sells the knives does, and it says the opposite of the folklore:
Angle of the cutting edge does not affect the ability to cut things.
Its figures are the surprise. Professional sashimi knives, it says, are often sharpened at about 35 to 45 degrees at the very edge, while general household knives are sometimes at about 25 to 30. The sharper feel of an acute edge is friction, not cutting ability, and the acute edge chips when it meets a pumpkin. I went through this in more detail in the piece on single and double bevels, where Tojiro describes a working edge as three angles, not one.
Tojiro Knife Gallery does name a hardest part, and it is not a number either:
研ぎで一番重要なことは研ぐときの包丁を寝かせる角度。「ハンドルを握る手首をいかに固定するか」という点が重要な部分です
The most important thing in sharpening is the angle at which you lay the knife down. What matters is how well you can hold the wrist that grips the handle still.
They sell a clip-on holder for exactly that, which tells you where they think beginners fail. Not in choosing the grit. In holding the same angle for two minutes.
If your knife is a single bevel, the back has its own rule
Aritsugu adds one warning for single bevel owners that is easy to ruin a knife with:
片刃の包丁の場合、裏面の凹みがなくなっていたら裏はいっぱい研いではダメだとか、使っている砥石の面は平らですか、とかね
For a single bevel knife, if the hollow on the back has gone, you must not keep sharpening the back. And is the face of the stone you are using flat?
The hollow on the back of a single bevel is ground in at the factory and is finite. Grind it away on a dished stone and the knife does not come back. If you own a yanagiba or a deba, read that sentence before your next session.
What I would buy
I do not buy knives or stones to test them, and I will not pretend otherwise. What follows rests on published specifications, on prices I checked myself, and on what three Japanese houses tell their own customers in Japanese.
If you own one Japanese knife and no stone: King 1000 grit Middle, $45.99 plus the Naniwa flattening stone, $27.99. That is $73.98 and it is the whole kit. 34mm of medium stone is the most sharpening per dollar on the list, it is heavy enough to sit still, and the second item is what keeps the first one honest.
If you will not soak a stone for ten minutes before cooking: Shapton Kuromaku 1000 grit, $53.99. Shapton ceramics need water on the surface, not a soak, which is the difference between sharpening tonight and sharpening one day. You are buying 15mm instead of 34mm for the convenience. Still add the flattening stone.
Buy the 8000 grit finisher third, if at all. A Tsukiji knife shop thinks 2000 grit finishes a kitchen knife. Get a flat 1000 grit working first, then decide whether you want polish.
Marked down: the combination stone as a beginner bargain. The Tojiro 1000 and 3000 with a stand at $44.99 is a decent object with a holder included, and the holder solves the sliding problem. But 24mm shared by two grits is not a lot of stone, and the second grit is the one the Japanese sources say you needed least.
Marked down: any guide that gets you to 6000 grit without telling you the stone goes hollow. Two of the four I read never mention it. Credit to Blade and Board, which gives it a section of its own and says plainly that whetstones dish over time and develop a concave hollow in the middle. Oishya mentions a flattener in passing, in half a sentence. Neither of the other two says a word.
Disclosure
I have four approved affiliate partnerships: two with knife retailers, Hocho-Knife and SharpEdge, and two with Japanese language-learning companies unrelated to knives. This page carries six tracked links to Hocho-Knife, pointing at four products. At about 10 percent, the King 1000 grit pays me about $4.60, the flattening stone about $2.80, the Shapton 1000 grit about $5.40 and the 8000 grit finisher about $9.10.
The item that pays me most is the one I told you to buy third. The item I recommend to every reader is the $27.99 flattening stone, which pays me the least of the four. The Tojiro combination stone is deliberately not a tracked link: Tojiro is one of seven brands Hocho-Knife excludes from its programme, so I could not be paid for it anyway, and marking it down costs me nothing. You should weigh that when reading the paragraph where I mark it down.
Tojiro is also a main source for this article, through the interview it gave to Naniwa, and Naniwa is the maker of the flattening stone I recommend. That interview is published by Naniwa on its own site, so treat the enthusiasm for Naniwa stones in it as advertising and the complaint about dished stones as the part that was not written to sell anything.
My full affiliate disclosure is here.
Sources used here, all public: the Naniwa interview series 砥石のこと, specifically the entries on Tsukiji Aritsugu and the Tojiro Knife Gallery; the Shapton corporate site for the flattening plate copy, the wear claim and the micron figures; the Hocho-Knife help pages on factory edge finishing and the store guidance on edge angle; the Hocho-Knife listings for every stone priced above, checked on 21 September 2026; and the four English grit guides named at the top, read the same day.