The frustrating answer first: on most automatic watches you cannot feel when the mainspring is full, and that is deliberate engineering rather than a shortcoming. So there are two reliable methods, one that takes ten seconds and one that takes two days. The ten-second method is to count crown turns — about 30 to 40 from a stopped watch. The two-day method is to wind it, leave it absolutely still, and time how long it runs before it stops, then compare that against the reserve figure the maker publishes for the caliber.
Why there is no hard stop at the crown
In a hand-wound watch the mainspring is hooked solidly to the inside of its barrel. Wind it to the end and it stops dead; keep forcing it and something breaks. That unambiguous stop is how owners of hand-wound watches know they are done.
An automatic cannot work that way, because the rotor on your wrist never stops trying to wind. Instead the outer end of the mainspring is held by a slipping bridle: a sprung strip that grips the barrel wall under normal tension and slips once the spring is fully wound, letting the excess go. That single component is why an automatic is mechanically impossible to overwind, by rotor or by a watch winder, and it is also why the crown keeps turning forever with roughly the same feel.
Some people describe a faint change in resistance as the bridle starts to slip. It is real but subtle, it varies by caliber, and it is not something to bet on. Counting is better.
Method 1: count the turns
From a fully stopped watch, hold the crown in its neutral position — pushed in, or in the first position on a screw-down crown once you have unscrewed it — and turn it in the winding direction:
- About 20 turns: the watch is running happily with enough reserve for a day. This is all you need to start a stopped watch and set it.
- About 30–40 turns: a full wind on a typical modern automatic. Past this point the bridle is slipping and you are achieving nothing.
- Beyond that: harmless but pointless. The thing to avoid is not over-winding, it is wearing out the crown threads and the stem by doing this every day on a screw-down crown. If a watch needs winding daily, wear it more or use a winder.
Two cautions. First, not every automatic hand-winds — older Seiko 5s and the original Orient Mako, for instance, had calibers that did not. If turning the crown produces no sound or resistance at all, check the caliber before you keep going; the winding guide covers how to tell. Second, wind the watch off the wrist. Winding it on your wrist pulls the stem sideways, and side load on a stem is a real way to damage one.
Method 2: time the reserve against the published figure
This is the definitive test, and it doubles as the best health check you can perform on an automatic without opening it. Wind the watch fully, set it, note the time, and put it somewhere it will not be touched — dial up, undisturbed, not on a winder. Then record when the seconds hand stops.
Now compare. The figures below are published by the makers for the calibers, not estimated by us:
- Seiko 4R35 / 4R36: approximately 41 hours, per Seiko’s own caliber specification (which also states 24 jewels, 21,600 vibrations an hour and a rate of +45 to −35 seconds a day worn on the wrist).
- Orient F6922: approximately 40 hours, per Orient’s published sheet for the Kamasu.
- Citizen 8210: 42 hours, per Citizen’s own listing for the Tsuyosa, which also states a hack feature.
Land within a few hours of the published number and you now know two things: your winding routine really does fill the spring, and the movement is in good order. Come in dramatically short — 20 hours against a published 41 — and the watch is telling you something. A short reserve points at the mainspring, the barrel or tired lubrication, and it is the most common honest reason to book a service.
The reason this matters: accuracy sags as the spring unwinds
A mainspring does not deliver even torque across its whole reserve. The last few hours before a watch stops are the least accurate hours it has, because the escapement is receiving the weakest push. This is why a watch that is nearly unwound every morning can look like a watch with an accuracy fault and is really a watch with a winding problem — and why the desk-bound week is the classic cause of both. Losing time and accuracy in general both start here.
It is also why “I wear it every day, why does it stop at the weekend?” has a boring answer: wearing a watch winds it in proportion to how much your wrist actually moves, and a quiet office day may add less than the day consumes. How many hours a day you need to wear one is the full version.
Three watches whose makers publish the figure to test against
If you want to run the reserve test, it helps to own a watch whose maker states a number. These three do — and that published figure is a far better reason to buy one than any claim about accuracy.
#1Best documented reserve
Seiko 5 Sports SRPD65
The default first automatic, and it earns it. Seiko's own 4R36 movement with day-date, hacking and hand-winding, a 100m case and a genuine in-house pedigree — for a price that makes it the reference point every other budget automatic is measured against.
Strengths
- +In-house Seiko 4R36 movement with hacking and hand-winding
- +Day-date, 100m water resistance and a see-through caseback onto the movement
- +The most cross-shopped, best-supported budget automatic there is
Trade-offs
- −The 4R36 is reliable, not accurate to chronometer standards
- −Hardlex crystal scratches more easily than sapphire
| Movement | Seiko 4R36 (automatic + hand-winding) |
|---|
| Stated accuracy | +45 / -35 sec/day |
|---|
| Power reserve | 41 hours |
|---|
| Case size | 42.5 mm |
|---|
| Water resistance | 100 m |
|---|
| Crystal | Hardlex mineral |
|---|
| Hacks / hand-winds | Yes / Yes |
|---|
Why this one. The 5 Sports line is the modern heir to the original Seiko 5: an honest, serviceable automatic with real heritage, which is exactly why it anchors so many of our lists. Seiko sells the SRPD65 head on more than one band; the listing linked here is the K4 variant, and the case, caliber and crystal figures above are Seiko's own for the reference rather than for a particular band.
Specs read from the product listing and the maker’s published data, on October 6, 2026. “Not published” means the maker does not state that figure.
#2Best in-house alternative
Orient Kamasu Diver (RA-AA0003R)
The value dive watch to beat. The Kamasu (Mako III) runs Orient's proven in-house F6922 — which hacks and hand-winds — behind a sapphire crystal, on a true 200m case. That sapphire is the main hardware upgrade over the older mineral-crystal Mako II, and it all lands for well under the Seiko it competes with.
Strengths
- +Sapphire crystal at a price where mineral is the norm
- +True 200m rating with a screw-down crown and solid bracelet
- +In-house F6922 movement that both hacks and hand-winds
Trade-offs
- −The in-house movement isn't chronometer-accurate
- −Roughly 42mm with a broad case — wears large on a small wrist
| Movement | Orient F6922 (automatic + hand-winding) |
|---|
| Stated accuracy | Not published |
|---|
| Power reserve | 40 hours |
|---|
| Case size | 41.88 mm |
|---|
| Water resistance | 200 m |
|---|
| Crystal | Sapphire |
|---|
| Hacks / hand-winds | Yes / Yes |
|---|
Why this one. The Kamasu's real upgrade over the Mako II is the sapphire crystal — both share the same F6922 movement, so both already hack and hand-wind. (The Orient diver that does neither is the original Mako I, on the older 46943 caliber.)
Specs read from the product listing and the maker’s published data, on September 15, 2026. “Not published” means the maker does not state that figure.
#3Longest published reserve here
Citizen Tsuyosa (NJ0)
The integrated-bracelet sports watch look, at a fraction of the watches that started the trend. A sunburst dial, a slim case and a comfortable bracelet make the Tsuyosa the affordable way to scratch that particular itch — with Citizen's own automatic movement inside.
Strengths
- +On-trend integrated bracelet and sunburst dial
- +Citizen in-house automatic movement
- +Wears slim and comfortable for an everyday watch
Trade-offs
- −50m water resistance — splash-proof, not a swimmer
- −Movement is reliable rather than high-accuracy
| Movement | Citizen 8210 (Miyota-based, automatic) |
|---|
| Stated accuracy | Not published |
|---|
| Power reserve | 42 hours |
|---|
| Case size | 40 mm |
|---|
| Water resistance | 50 m |
|---|
| Crystal | Sapphire |
|---|
| Hacks / hand-winds | Yes / Yes |
|---|
Why this one. The Tsuyosa rode the integrated-bracelet wave and became a genuine hit by delivering the silhouette honestly and cheaply. One caveat worth knowing: the earliest units did not hack, while current production does — Citizen doesn't publish a daily accuracy figure either way.
Specs read from the product listing and the maker’s published data, on September 30, 2026. “Not published” means the maker does not state that figure.
All three hand-wind, which is what makes the test possible at all, and all three hack so you can set them precisely against a reference clock. If you are shopping rather than diagnosing, the wider lists are best Japanese automatics and best entry-level automatics.