Hone at roughly the angle the knife was sharpened to: about 20 degrees per side for Western and German knives, about 15 for Japanese. That is the answer, and it is the same pair of numbers as the sharpening angle — which is exactly why so much writing treats the two questions as one. They are not one question, because the tolerance is completely different.
On a stone, angle error accumulates. Every stroke at 23 degrees on a knife ground to 20 grinds a new, wider bevel, and a hundred strokes later you own a duller knife with a thicker edge. That is why stone work rewards obsessive angle control. Honing removes essentially no metal, so nothing accumulates. Five degrees off on a rod does not reshape anything — it just fails to do the job, or does part of it. That is a far more forgiving failure, and it changes what you should worry about.
Which way you miss is what matters
There are two ways to be wrong, and they are not symmetrical.
Too shallow— the blade lying too flat against the rod — is the harmless miss and by far the more common one. The rod rides on the bevel behind the apex and never contacts the rolled part at all. Nothing is damaged; nothing is fixed either. This is the honest explanation for most of the "my honing rod does nothing" complaints on the internet: not a bad rod, not a worn edge, just an angle that never reached the apex.
Too steep— the spine lifted too high — is the miss that costs you. The rod now contacts only the very apex at a wider angle than the bevel, and light as the pressure is, repeated steep passes round the apex over. The knife gets progressively less crisp and you hone harder to compensate, which makes it worse. If you are going to be wrong, be wrong shallow.
The practical consequence is that you should start each stroke deliberately shallow, then lift the spine slightly until you feel the rod bite rather than skate. That felt change is the apex making contact. It is a better guide than any number.
What 15 and 20 degrees actually look like
Nobody can eyeball a degree, so use references instead of numbers.
The halving method.Hold the blade flat against the rod — that is 0 degrees. Lift it to perpendicular, 90. Halve it to 45, halve again to about 22. That last position is close enough to 20 for honing, and you can find it in two seconds without thinking about it.
The matchbook trick. A matchbook laid flat under the spine of an 8-inch blade puts the edge at roughly 20 degrees. It is the reference professional kitchens have used for decades and it is accurate enough for a job with this much tolerance.
For 15 degrees, halve the 20-degree position toward the rod by a small amount rather than trying to find 15 from scratch — you will land between 15 and 18, and on a Japanese edge that is the correct side to miss on.
Why the honing angle can be slightly wider than the bevel
A common and reasonable question: should you hone at exactly the sharpening angle, or a degree or two above it? A rod contacting slightly above the bevel angle reaches the apex reliably, which is what you want, and because it removes essentially no metal it is not creating a secondary bevel the way a stone at the same angle would. That is the argument for a hair steeper. The counter-argument is that "a hair" is not a quantity anyone can hold, and steep is the destructive miss. Our reading: aim at the bevel angle, accept that you will land a degree or two either side, and let the felt bite tell you when you have contact.
Pressure and count
Light. The weight of the knife plus a little is enough — if the rod rings loudly and your forearm is working, you are pressing far too hard, and on a ceramic rod that is how the surface gets loaded with steel. Five to ten alternating passes per side handles a normal roll. If twenty passes have not brought the bite back, stop: the edge is worn rather than rolled and needs a stone, and continuing just rounds the apex further.
Serrated knives, and the one angle rule that changes
A serrated edge is ground on one side only, so there is no alternating stroke and no symmetrical angle to hold. A tapered ceramic rod goes into each scallop on the ground side and the flat back is deburred afterwards. Do not use a normal honing motion on serrations and do not alternate sides — how to sharpen a serrated knife covers the whole procedure.
Where the sharpening angle question lives
Everything above is about honing. If you are setting an angle on a stone — where error genuinely accumulates and the 15-versus-20 decision permanently changes the knife — that is a different page: knife sharpening angles covers the per-side geometry, why a fixed 20-degree pull-through quietly re-profiles a Japanese edge, and how to decide which angle a given knife should carry.
How we picked
We did not lab-test these knives
Everyone in this category says they tested twenty knives in a professional kitchen. We have not hands-on tested any of these, and we say so. What we did instead: compiled the published specifications — steel type and hardness (HRC), edge angle, blade geometry, weight and build — applied established sharpening standards, ran the value math (spec quality ÷ live price), and scored each product against a published rubric. The scores are judgments from documented research — not measurements we took, because we do not have a test lab and we will not pretend we do. Where a number came from someone else's work, we name them in Sources.
