Watch Accuracy Testing Methods: How to Check if Your Timepiece is Keeping Good Time

So I got this beautiful automatic watch for my birthday last year — a Seiko Presage that I’d been eyeing forever. After a couple of months, I started noticing something weird. It seemed to be running a bit slow. Like, I’d set it perfectly in the morning, and by evening, it would be a few minutes behind. Was I imagining things? Was my beloved new watch defective? How could I even tell for sure?

If you’ve ever wondered whether your watch is actually keeping good time or if it’s just in your head, you’re definitely not alone. Watch accuracy is one of those things that watch enthusiasts (we’re a weird bunch, let’s be honest) can get really obsessive about. But figuring out exactly how accurate your watch is running isn’t always straightforward.

After falling down a pretty deep rabbit hole of research and experimentation, I’ve put together this guide on different ways to test your watch’s accuracy — from super simple methods that cost absolutely nothing to the fancier approaches if you want to get really precise. Let’s get into it!

What is Watch Accuracy Anyway?

Before we dive into testing methods, we should probably talk about what we’re actually measuring and what’s considered “good” accuracy.

Watch accuracy is typically measured in seconds per day (spd) — how many seconds your watch gains or loses over a 24-hour period. A watch that runs “fast” gains time (the watch says it’s 10:05 when it’s actually 10:00). A watch that runs “slow” loses time (the watch says it’s 9:55 when it’s actually 10:00).

As for what’s considered “good” accuracy, it really depends on the type of watch:

• Quartz watches: Typically accurate to ±15-20 seconds per month (or about ±0.5-0.7 seconds per day)
• Standard mechanical watches: Usually around ±10-30 seconds per day
• COSC-certified chronometers: Between -4 and +6 seconds per day
• Grand Seiko/high-end standards: Often around ±5 seconds per day or better

My Seiko’s movement is supposed to be accurate to about +25/-15 seconds per day, which I learned is pretty standard for watches in its price range. Honestly, I was a bit shocked when I first found out mechanical watches could be off by that much and still be considered “normal.” Coming from a quartz watch that barely drifted a minute over several months, this was a whole new world!

Basic Methods: No Special Equipment Required

Let’s start with the easiest ways to check your watch’s accuracy that don’t require buying anything special.

The Time.is Method

This is my go-to method because it’s so darn simple:

1. Set your watch precisely using the website Time.is (it syncs with atomic clocks)
2. Wait a consistent period (24 hours is standard, but you can do longer)
3. Check the difference between your watch and Time.is again
4. Calculate the seconds per day rate

Let’s say after exactly 24 hours, your watch shows it’s 10:30:15, but Time.is shows 10:30:27. That means your watch has lost 12 seconds in a day, so it’s running at -12 seconds per day.

Pro tip: Take a photo of your watch next to the time.is website when you start. This helps eliminate any reaction time issues when you check again later.

The Smartphone App Method

There are several free apps designed specifically for watch accuracy tracking:

• Watch Tracker (iOS)
• Atomic Clock & Watch Accuracy (Android)
• WatchCheck (Android)

These apps let you log the deviation between your watch and the reference time over multiple days, which is super helpful because—and this surprised me—a watch’s accuracy isn’t always consistent. My Seiko, for example, runs a bit faster when it’s fully wound than when the power reserve is low.

The apps typically calculate the average daily rate over time, giving you a much more reliable picture than a single measurement.

The Observation Method

If you’re not concerned with exact numbers and just want to know if your watch is “reasonably accurate,” this method works fine:

1. Set your watch to the exact time
2. Go about your normal life, wearing the watch as you typically would
3. Check after a week to see how far off it is
4. Divide by 7 to get the average daily rate

When I did this with my Seiko, I found it was about 90 seconds slow after a week, which works out to roughly -13 seconds per day. Not perfect, but well within the manufacturer’s specifications.

Intermediate Methods: Minimal Investment

Ready to get a bit more precise? These methods require some minor purchases but give you better results.

The Timegrapher App

Apps like “Hairspring” for iOS or “Watch Accuracy Meter” for Android turn your smartphone’s microphone into a basic timegrapher—a device that listens to the ticking of your watch and calculates its accuracy.

These apps work by:

1. Placing your watch near your phone’s microphone
2. The app recording and analyzing the sound of the watch ticking
3. Calculating the rate based on the timing between ticks

While not as accurate as dedicated equipment, these apps can give you a reasonable approximation. I tried Hairspring with my watch, and it showed around -10 seconds per day, which was pretty close to my manual measurements.

The catch: These apps work best with watches that tick loudly enough for your phone’s microphone to pick up. My dive watch works great, but my dress watch with its quieter movement? Not so much.

Reference Watch Method

If you have a highly accurate watch (like a radio-controlled G-Shock or an Apple Watch), you can use it as your reference:

1. Set your mechanical watch to exactly match your reference watch
2. After several days, compare the two
3. Calculate the daily rate based on the difference

I actually did this when I first got suspicious about my Seiko’s accuracy. I set it to match my G-Shock (which syncs with atomic time), and checked again after three days. The Seiko was about 40 seconds behind, suggesting around -13 seconds per day.

Position Variation Testing

This one blew my mind when I first learned about it—a mechanical watch’s accuracy can vary significantly depending on the position it’s in! This happens because gravity affects the movement differently depending on orientation.

To test position variation:

1. Use one of the methods above to measure accuracy with the watch dial-up
2. Repeat with the watch in different positions (dial down, crown up, crown down, etc.)
3. Note the differences between positions

When I tried this with my Seiko, I discovered it runs about 5 seconds slower per day when positioned crown-up overnight compared to dial-up. This is actually useful knowledge—if I know my watch runs a bit slow, I can leave it crown-down when I’m not wearing it to help compensate!

Advanced Methods: For the Really Committed

If you’re getting seriously into watch accuracy, these methods provide the most precise measurements.

Professional Timegrapher

A real timegrapher is the tool professionals use, and home versions start around $150-300 (like the Weishi 1000 or Timegrapher 1900).

These machines:

• Precisely measure the timing of each tick
• Calculate accuracy in seconds per day
• Show beat error and amplitude (other important metrics)
• Allow testing in all positions

The advantage is incredibly precise, instant readings without waiting days to see results. They also help diagnose specific issues with the movement.

I haven’t invested in one of these yet, but a watch enthusiast friend let me test my collection on his. My Seiko showed -11 seconds per day in the dial-up position, which matched pretty closely with my manual observations.

Multi-day Position Testing

The most thorough home testing method combines extended observation with position testing:

1. Test your watch in each position (dial up, dial down, crown up, etc.)
2. Leave it in each position for a full 24 hours
3. Record the daily rate in each position
4. Calculate the average across all positions

This gives you a comprehensive understanding of your watch’s performance in actual use, since while wearing it, the watch naturally moves through many positions.

Regulation or Service

If your testing shows that your watch is running outside acceptable parameters, you have options:

• Regulation: A relatively simple adjustment of the regulator to speed up or slow down the watch
• Full service: A complete disassembly, cleaning, and adjustment

After tracking my Seiko for a few months and consistently finding it running around -13 seconds per day, I decided to have it regulated. My watchmaker adjusted it to about +2 seconds per day, which I actually prefer—I’d rather my watch be slightly ahead than behind.

Factors That Affect Mechanical Watch Accuracy

Through my testing adventures, I’ve learned that several factors can affect how accurately a mechanical watch runs:

Movement Quality

Simply put, higher-quality movements tend to maintain better accuracy. This is why watches with premium movements from brands like Grand Seiko, Rolex, or Omega typically outperform more affordable options like my Seiko. They use better materials, more precise manufacturing, and more careful adjustment from the factory.

Age and Service History

A watch that hasn’t been serviced in 10+ years will almost certainly not keep time as well as it did when new. The oils dry up, parts wear, and performance degrades. After servicing, most watches return to something close to their original accuracy.

Wearing Patterns

This one surprised me—how often and how long you wear your watch affects its accuracy:

• Automatic watches typically perform more consistently when worn regularly
• Hand-wound watches often show different rates at different states of wind
• How active you are can impact how fully wound an automatic stays throughout the day

I noticed my Seiko runs most accurately when I wear it daily for at least 12 hours with normal activity. Weekend desk duty? It tends to slow down a bit more.

Environmental Factors

Temperature, magnetism, and shocks can all affect accuracy:

• Temperature: Most mechanical watches are regulated for average wrist temperature (~98°F/37°C)
• Magnetism: Everyday electronic devices can magnetize watches, causing significant deviation
• Shocks: Hard bumps can temporarily or permanently affect timekeeping

I once accidentally left my watch near my laptop’s speaker overnight, and suddenly it was running +45 seconds per day instead of -13! A quick demagnetization at my local watchmaker solved the problem completely.

Comparing Accuracy Across Watch Types

Different watch technologies have inherent accuracy capabilities:

Quartz Watches

Standard quartz watches are typically accurate to ±15 seconds per month. They achieve this by using a quartz crystal that vibrates at a very precise frequency when electricity is applied.

High-accuracy quartz watches like Citizen’s Chronomaster or Bulova’s Precisionist can achieve ±5 seconds per year by using special techniques like thermocompensation.

Mechanical Watches

Standard mechanical watches usually range from ±10-30 seconds per day. They rely on a balance wheel and hairspring oscillating at typically 28,800 vibrations per hour.

COSC-certified chronometers are tested to run between -4 and +6 seconds per day in various positions and temperatures.

Spring Drive (Seiko/Grand Seiko)

This hybrid technology uses a conventional mainspring for power but regulates it with a quartz oscillator. Spring Drive watches typically achieve ±1 second per day accuracy—far better than standard mechanical movements but not quite at quartz levels.

Radio-Controlled/GPS Watches

These watches sync with external time references (radio signals or GPS satellites) and are therefore accurate to the second when synchronized. Between syncs, they rely on standard quartz accuracy.

Is Perfect Accuracy Actually Important?

After all my testing and tweaking, I’ve come to a somewhat surprising conclusion: perfect accuracy isn’t always the most important thing for a mechanical watch.

Don’t get me wrong—it’s satisfying when a watch keeps excellent time. But I’ve realized that expecting mechanical perfection kind of misses the point of these tiny machines. They’re remarkable precisely because they can achieve such good timekeeping using only springs and gears.

My watchmaker shared a perspective that stuck with me: “Mechanical watches aren’t about perfect timekeeping anymore. If that’s all you want, use your phone. They’re about the art of watchmaking, the engineering achievement, and the connection to something that doesn’t need a battery to live.”

That said, there’s a reasonable expectation of performance. A watch running several minutes off per day probably needs attention. But whether your watch runs at +2 or -4 seconds per day? That’s just part of the character of mechanical timekeeping.

My Personal Accuracy Testing Routine

After trying all these methods, I’ve settled into a simple routine for my collection:

1. I use the WatchTracker app to spot-check new watches over their first week
2. If performance seems questionable, I’ll do a more thorough 5-position test
3. For my regularly worn watches, I check accuracy about every 6 months
4. Any watch running outside of ±20 seconds per day gets evaluated for service

This system gives me peace of mind without turning into an obsession. Because let’s face it, it’s easy to go down the accuracy rabbit hole and forget to actually enjoy your watches!

Final Thoughts: Finding Your Accuracy Comfort Zone

The bottom line? How accurately you need your watch to run is a personal decision. Some collectors are perfectly content with ±30 seconds per day. Others won’t settle for anything more than ±2.

I’ve found my comfort zone is about ±10 seconds per day for watches I wear regularly. That means they’re never more than a minute off between weekly adjustments, which works for me.

Whatever method you use to test your watches, remember that mechanical watches are miniature marvels that somehow keep reasonable time using technology essentially hundreds of years old. A little variation isn’t just acceptable—it’s part of their charm!

How accurate is your favorite watch? Have you tried testing it with any of these methods? I’d love to hear about your experiences in the comments!

SEE also

Rolex
Omega
Patek Philippe
Audemars Piguet
TAG Heuer
Seiko
Longines
Tissot
Casio
Citizen