Not all freezes are the same problem, and treating them as one is why people spend a weekend reinstalling Windows only to have it lock up again on Tuesday. A half-second stutter every few minutes, a hard lock that needs the power button, a freeze that ends in a blue screen, and a freeze that only happens in QuickBooks are four different faults with four different causes.

So start by characterizing what you actually have. Then go to Event Viewer, because Windows almost always wrote down what happened.

Read the logs before you change anything

Press Windows + R, type eventvwr.msc, press Enter. In the left pane expand Windows Logs > System. On the right, click Filter Current Log, tick Critical and Error, and click OK.

Now find the timestamp of your last freeze and look at what’s logged around it. A few entries matter more than the rest:

• Kernel-Power, Event ID 41. “The system has rebooted without cleanly shutting down first.” This confirms a hard lock or power loss, but on its own it doesn’t say why. Look at what’s immediately before it.
• Event ID 7 or 51, source disk. “A parity error was detected” or “An error was detected on device during a paging operation.” Strong signal of a failing drive or cable.
• Event ID 129, source storahci. “Reset to device, DeviceRaidPort0, was issued.” The storage controller stopped responding — classic freeze cause.
• Event ID 4101, source Display. “Display driver nvlddmkm stopped responding and has successfully recovered.” A GPU driver problem.
• Event ID 1001, BugCheck. Records the blue screen stop code — the single most useful line if you got a BSOD.

Also check Windows Logs > Application for Event ID 1000 (application crash) if only one program locks up. It names the faulting module, which frequently points straight at an outdated printer driver or add-in.

Ten minutes here saves hours of guessing.

Freeze pattern to probable cause

The five-minute checks

Free disk space

Windows needs headroom for the page file, temp files, and updates. Under about 10% free on the system drive, performance falls apart and freezes start. Under 5 GB free, expect trouble.

Go to Settings > System > Storage, turn on Storage Sense, and run Cleanup recommendations. Old Windows update files and a leftover Windows.old folder often account for 15–30 GB by themselves.

Memory pressure

Open Task Manager (Ctrl+Shift+Esc) and check the Performance tab. If memory sits above 85% during normal work, you’re paging to disk constantly and that feels exactly like freezing.

8 GB is now the floor for a Windows 11 machine and it’s a tight floor if someone runs Chrome with thirty tabs alongside Excel and Teams. 16 GB is the practical standard for office work. RAM is one of the cheapest fixes available.

Disk at 100%

In Task Manager’s Processes tab, sort by the Disk column. If it pins at 100% while responsiveness dies, look at what’s on top. Common culprits: Windows Search indexing after a big file move, an antivirus full scan running during business hours, Windows Update, OneDrive syncing thousands of files, or a mechanical hard drive that simply can’t keep up with a modern OS.

That last one is worth stating plainly: if the machine still has a spinning hard drive, replacing it with an SSD fixes more freezing complaints than any other single change. It’s a bigger improvement than a new CPU.

Testing the RAM properly

Bad memory causes freezes with no pattern at all — that’s its signature. Same task, works fine ninety times, locks the machine on the ninety-first.

Windows has a built-in test. Press Windows + R, type mdsched.exe, choose “Restart now and check for problems.” It runs before Windows loads and takes 15–30 minutes. Results appear afterward in Event Viewer under Windows Logs > System, source MemoryDiagnostics-Results.

Be aware it’s a fairly shallow test. It catches obvious failures and misses intermittent ones. For a machine that freezes randomly and passes the Windows test, run MemTest86 from a USB stick and let it complete at least four full passes — usually overnight. A single error is a failed module.

If you have two sticks, you can also test by elimination: run with one installed for a few days, then swap. Slower, but free and conclusive.

Checking the drive

Open an elevated command prompt (right-click Start, choose Terminal (Admin)) and run:

• wmic diskdrive get status — a crude pass/fail. “Pred Fail” means back up now.
• chkdsk C: /scan — scans without needing a reboot or locking the volume.
• sfc /scannow — repairs corrupted Windows system files.
• DISM /Online /Cleanup-Image /RestoreHealth — fixes the component store when sfc can’t.

The built-in status check is not very sensitive. A free tool like CrystalDiskInfo shows actual SMART attributes, and for SSDs the ones to watch are reallocated sector count, media wear indicator, and total bytes written. For mechanical drives, any nonzero reallocated or pending sector count means the drive is failing.

Heat and thermal throttling

If freezes happen during demanding work — exporting video, running large reports, a long Excel recalculation — and the fans get loud right before, temperature is the first thing to check.

Install HWiNFO or Core Temp and watch CPU temperature during the task. Modern Intel and AMD chips throttle around 95–100 °C and will shut down or lock beyond that. Anything sustained above 90 °C on a laptop means the cooling isn’t keeping up.

The usual causes are dust packed into the heatsink fins, thermal paste dried out after four or five years, a fan that’s seized or spinning slowly, or a laptop being used on a couch cushion that blocks the intake. Compressed air through the vents helps in the short term; a proper teardown and repaste is the real fix on an older machine.

In an office, look at where machines physically sit. A tower shoved into a cabinet with no airflow, or a laptop docked and closed with the vents against a wall, will overheat predictably at 3 p.m. every day.

Drivers, especially graphics

If the picture freezes but you can still hear audio, or the screen goes black for a second and returns with a “display driver stopped responding” notice, that’s the GPU driver.

Don’t rely on Windows Update for graphics drivers. Get them from Intel, NVIDIA, or AMD directly. For a laptop, the manufacturer’s version (Lenovo, Dell, HP) is often better tuned to that specific hardware.

If the freezing started right after a driver update, roll it back: open Device Manager, expand Display adapters, right-click the GPU, Properties > Driver > Roll Back Driver. If that button is greyed out, uninstall the driver with “Delete the driver software for this device” ticked and install the previous version manually.

Chipset and storage drivers matter too. A generic Microsoft AHCI or NVMe driver in place of the vendor one causes exactly the kind of stalls that show up as storahci Event ID 129.

Isolate software with a clean boot

If you suspect software rather than hardware, a clean boot narrows it fast. Press Windows + R, type msconfig, go to the Services tab, tick Hide all Microsoft services, click Disable all. Then in Task Manager’s Startup apps tab, disable everything. Reboot.

If the freezing stops, re-enable items in halves until it returns. Frequent offenders in small offices: two antivirus products fighting each other, ancient printer or scanner utilities, backup agents from a service nobody uses anymore, and vendor “helper” applications that came with a monitor or webcam.

Running more than one real-time antivirus is worth calling out specifically. It’s a reliable way to produce freezes, and it happens constantly at businesses that bought a security product without removing the previous one.

Malware

Cryptominers and other unwanted software pin the CPU and make everything feel frozen. Look in Task Manager for a process using heavy CPU with a name you don’t recognize, especially one that drops to zero the moment you open Task Manager.

Run a full offline scan: Windows Security > Virus & threat protection > Scan options > Microsoft Defender Antivirus (offline scan). It reboots and scans before Windows loads, which catches things that hide from a normal scan. Follow up with a second-opinion scanner such as Malwarebytes.

If one office machine is confirmed infected, check the rest. Malware that arrived by email or a shared drive rarely stops at one PC.

The power supply nobody suspects

On desktops, a failing PSU produces freezes and hard locks that look like anything else. The tell is that lockups get worse under load and the machine sometimes powers off entirely with no blue screen or event log entry beyond Kernel-Power 41.

Cheap power supplies degrade after three to five years. If you’ve tested RAM, drive, temperatures, and drivers and found nothing, and the machine has a bargain PSU or is plugged into an overloaded strip, that’s the remaining candidate. Swapping in a known-good unit is the only practical test.

Related: if several machines in the same room lock up during the same part of the day, look at the building’s electrical situation before the computers. A shared circuit with a space heater or a compressor on it can cause voltage sags. A basic UPS on each workstation is cheap insurance and eliminates the variable.

Common questions

Why does my PC keep freezing but there’s no blue screen?

A blue screen means Windows detected a fault and stopped deliberately. A silent hard lock means it never got the chance — the hardware stopped responding first. That pattern points more toward memory, storage controller, power delivery, or heat than toward a software bug. Check Event Viewer for Kernel-Power 41 and look at whatever was logged just before it.

How long should I run a memory test?

The built-in Windows Memory Diagnostic takes 15–30 minutes and catches only clear failures. For intermittent freezing, run MemTest86 for at least four complete passes, which usually means leaving it overnight. Intermittent memory faults can take hours to surface, so a short clean run doesn’t prove much.

Will reinstalling Windows fix freezing?

Only if the cause is software — driver conflicts, corrupted system files, or accumulated junk. If the cause is failing RAM, a dying SSD, overheating, or a weak power supply, a fresh install fixes nothing and costs you a day. Rule out hardware first. That’s exactly why the diagnostics come before the reinstall.

One program freezes but the rest of the computer is fine. Is that hardware?

Almost never. A single misbehaving application points to that app — an add-in, a corrupt data file, a stale cache, or a network path it can’t reach. For Office, start it in safe mode by holding Ctrl while launching, or disable COM add-ins under File > Options > Add-ins. If the app opens a file from a network share, test with a local copy to see whether the network is the real bottleneck.

Several office PCs started freezing around the same time. What connects them?

Look for the shared factor rather than diagnosing each machine. Usual answers: a Windows update or driver pushed to all of them, a newly deployed security agent, a failing server or NAS that workstations wait on, an antivirus scan scheduled during business hours, or an electrical problem in the building. Check what changed in the week before the complaints started.