Your computer runs slow. Programs freeze. You hear strange noises from the case. A PC health check identifies exactly what’s failing before you lose data or face expensive repairs. This guide shows you how to diagnose hardware failures, measure system performance, and fix problems yourself.
What Is a PC Health Check?
A PC health check examines your computer’s critical components to find failures before they cause crashes or data loss. You’re checking:
- Storage drive condition and lifespan
- CPU and GPU operating temperatures
- Memory integrity and errors
- System file corruption
- Power supply stability
- Component performance benchmarks
Professional repair shops charge $50-100 for diagnostics. You can do the same checks yourself in under an hour using free tools.
Why Regular Health Checks Matter
Hard drives fail. RAM develops errors. Thermal paste dries out. These problems start small and get worse.
I’ve seen thousands of repair cases. Most people lose data because they ignored warning signs for months. A failing drive shows symptoms 2-3 months before complete failure. Check your system monthly and you’ll catch problems early.
Regular checks also reveal performance bottlenecks. You might think you need a new computer when you actually just need more RAM or an SSD upgrade.
Step 1: Check Storage Drive Health
Storage failure causes 90% of data loss cases. Check your drives first.
For Hard Drives (HDD)
Hard drives use spinning platters. They fail mechanically. Signs include clicking sounds, slow file access, and disappeared files.
Check SMART data with CrystalDiskInfo:
- Download CrystalDiskInfo and install
- Open the program
- Look at the health status at the top
Health indicators:
| Status | Meaning | Action |
|---|---|---|
| Good (Blue) | Drive is healthy | Continue normal use |
| Caution (Yellow) | Drive is degrading | Backup data immediately, plan replacement |
| Bad (Red) | Drive is failing | Stop using, recover data now |
Critical SMART attributes to check:
- Reallocated Sectors Count: Shows damaged sectors the drive has marked as unusable. Any number above 0 means the drive is degrading. Above 50 means replace immediately.
- Current Pending Sectors: Sectors waiting to be remapped. Above 0 indicates active problems.
- Uncorrectable Sector Count: Data that can’t be recovered. Any number here is serious.
Temperature: HDDs should run at 25-45°C. Above 50°C shortens lifespan significantly.
For Solid State Drives (SSD)
SSDs fail differently than HDDs. They have limited write cycles. Modern SSDs last 5-10 years with normal use.
CrystalDiskInfo works for SSDs too. Additional checks:
Total Bytes Written: Shows how much data you’ve written to the drive. Compare this to the manufacturer’s TBW (terabytes written) rating. If you’re at 80% of rated TBW, plan replacement within a year.
Wear Leveling Count: Shows remaining lifespan. Most SSDs start at 100 and count down. Below 10 means the drive is near end-of-life.
Windows Built-in Drive Check
Windows includes basic drive diagnostics:
- Open Command Prompt as administrator
- Type:
wmic diskdrive get status - Press Enter
“OK” means Windows sees no problems. This check is superficial. Always use CrystalDiskInfo for real health data.
Run CHKDSK for file system errors:
- Open Command Prompt as administrator
- Type:
chkdsk C: /f /r /x - Type Y to schedule on restart
- Restart computer
This scans for bad sectors and fixes file system corruption. Run it quarterly.
Step 2: Monitor Temperature and Cooling
Heat kills computers. CPUs throttle performance at 85°C and can fail at 100°C. GPUs throttle at 90°C.
Most performance problems are actually overheating problems. Your system slows itself down to prevent damage.
Install Temperature Monitoring
Download HWiNFO for detailed sensor monitoring. It’s free and shows every temperature sensor in your system.
Setup:
- Install HWiNFO64 (not the 32-bit version)
- Launch with “Sensors-only” mode
- Look for these critical temps:
Normal temperature ranges:
| Component | Idle | Under Load | Critical |
|---|---|---|---|
| CPU | 30-50°C | 60-80°C | 85°C+ |
| GPU | 30-45°C | 65-85°C | 90°C+ |
| Motherboard | 30-45°C | 40-55°C | 70°C+ |
| M.2 SSD | 30-50°C | 50-70°C | 80°C+ |
Test under load:
- Open HWiNFO sensors window
- Run a demanding game or program for 20 minutes
- Watch maximum temperatures
If temps exceed critical thresholds, you have cooling problems.
Fix Overheating Issues
Dust buildup is the #1 cause. Open your case and check fans and heatsinks. Dust acts as insulation.
Cleaning process:
- Unplug computer completely
- Take case outside or to a garage
- Use compressed air to blow dust from fans, heatsinks, and filters
- Hold fans still while spraying so they don’t overspin
- Clean dust filters separately with water, let dry completely
Do this every 6 months minimum. Every 3 months if you have pets or carpet.
Thermal paste degradation: The paste between your CPU and cooler dries out after 3-4 years. This raises temps by 10-20°C.
Replacing thermal paste requires removing the CPU cooler. If you’re not comfortable doing this, it costs $30-50 at a repair shop.
Poor case airflow: Your case needs intake fans at the front and exhaust fans at the rear/top. Hot air rises naturally. Help it along.
Ideal setup:
- 2-3 intake fans (front/bottom)
- 1-2 exhaust fans (rear/top)
- Positive air pressure (slightly more intake than exhaust)
This keeps dust out and temperatures down.
Step 3: Test RAM for Errors
Bad RAM causes random crashes, blue screens, corrupted files, and weird behavior. Programs might close unexpectedly. Windows might freeze.
RAM errors are invisible until they cause damage. Test your memory regularly.
Windows Memory Diagnostic
Built into Windows and works well:
- Press Windows + R
- Type:
mdsched.exe - Choose “Restart now and check”
- Computer restarts and tests memory (20-30 minutes)
- Windows shows results after reboot
If errors appear: Your RAM is failing. Identify which stick is bad by testing one at a time. Replace faulty modules immediately.
MemTest86 (More Thorough)
MemTest86 is more comprehensive than Windows Memory Diagnostic. It catches errors Windows misses.
- Download MemTest86
- Create a bootable USB drive
- Boot from USB
- Let it run for at least 4 full passes (takes several hours)
Run MemTest86 if you have mysterious crashes but Windows Memory Diagnostic found nothing.
Zero errors is the only acceptable result. One error means bad RAM.
Step 4: Check System File Integrity
Windows system files can corrupt from bad shutdowns, malware, or drive errors. Corrupted files cause crashes, error messages, and weird behavior.
System File Checker (SFC)
SFC scans and repairs corrupted Windows files:
- Open Command Prompt as administrator
- Type:
sfc /scannow - Wait 15-30 minutes
- Read results
If SFC can’t fix files, run DISM first:
- Type:
DISM /Online /Cleanup-Image /RestoreHealth - Wait 20-40 minutes
- Run
sfc /scannowagain
This repairs Windows from Microsoft’s servers. Requires internet connection.
Run SFC every 3 months or after crashes/power failures.
Step 5: Scan for Malware and Threats
Malware causes slowdowns, crashes, high CPU usage, and privacy breaches. Windows Defender is decent but not perfect.
Run Multiple Scans
- Windows Defender: Already installed. Open Windows Security > Virus & threat protection > Scan options > Full scan
- Malwarebytes: Download free version from Malwarebytes, run full scan
- HitmanPro: Download from Sophos, excellent at finding rootkits
Different scanners catch different threats. Run all three if your system acts infected.
Signs of malware:
- Browser redirects to strange sites
- New toolbars or programs you didn’t install
- Slow performance despite low CPU usage in Task Manager
- Popup ads when browser is closed
- Disabled antivirus or Windows Update
After removal: Change all passwords from a different device. Malware often includes keyloggers.
Step 6: Performance Benchmark
Benchmarks show if components perform as expected. A failing CPU might run at 50% of normal speed. You won’t notice until you compare.
UserBenchmark Quick Test
UserBenchmark tests all major components in 5 minutes:
- Download and run UserBenchmark
- Close all programs first
- Let it complete testing
- View results online
Results show percentile rankings. Your component should score near the average for its model. If your CPU ranks in the bottom 10%, something is wrong (usually thermal throttling or power limits).
Individual Component Testing
CPU: Use Cinebench R23 for CPU-only stress testing. Compare your score to published reviews of your CPU model.
GPU: Use 3DMark Basic Edition (free on Steam) for graphics testing. Again, compare to reviews.
Storage: CrystalDiskMark tests read/write speeds. SSDs should hit their rated speeds. If your 500MB/s SSD only gets 200MB/s, it’s failing or connected wrong.
Step 7: Check Power Supply Health
Failing power supplies cause random shutdowns, restarts, and component damage. They’re often overlooked.
Signs of PSU failure:
- Random shutdowns under load
- System won’t power on reliably
- Burning smell from computer
- Coil whine or buzzing sounds
- Crashes when using GPU heavily
No software accurately tests PSUs. You need a multimeter or PSU tester. If you suspect PSU failure, test with a known good unit or take it to a shop.
PSU age: Most power supplies last 5-7 years. After that, capacitors degrade. If your PSU is over 7 years old and you have stability problems, replace it.
Creating a Maintenance Schedule
Don’t wait for problems. Check regularly.
Monthly:
- Check drive health (CrystalDiskInfo)
- Monitor temperatures under load (HWiNFO)
- Quick malware scan (Windows Defender)
- Visual dust inspection
Quarterly:
- Run SFC system file scan
- Full malware scan (Malwarebytes)
- Run CHKDSK
- Check all SMART attributes
Every 6 months:
- RAM test (Windows Memory Diagnostic)
- Performance benchmark
- Physical case cleaning
- Check all fan operation
Yearly:
- MemTest86 thorough RAM test
- Replace case fans if noisy or slow
- Verify backup systems work
Every 3-4 years:
- Replace CPU thermal paste
- Consider PSU replacement
- Evaluate upgrade needs
Set calendar reminders. Most people only check when problems appear. By then you might have lost data.
When to Repair vs Replace
Run the numbers before repairing.
Repair makes sense when:
- Single component failure (RAM, drive, fan)
- Computer is under 5 years old
- Repair cost is under $200
- Performance meets your needs otherwise
Replace instead when:
- Multiple components failing
- Computer over 7-8 years old
- Repair costs exceed 50% of new computer price
- You need significantly better performance
Example: If your 6-year-old computer needs a new motherboard ($150), RAM ($80), and power supply ($70), that’s $300. A new budget PC costs $500. Replace it.
But if your 3-year-old gaming PC just needs a new SSD ($80), repair it.
Conclusion
A PC health check takes 60 minutes and prevents data loss, expensive repairs, and downtime.
Check storage drives monthly with CrystalDiskInfo. Monitor temperatures when gaming or working. Test RAM twice yearly. Scan for malware monthly. Clean dust every 6 months.
Most computer problems show warning signs weeks or months early. Regular health checks catch them before they cause damage.
Start with the storage check. Drive failure causes irreversible data loss. Everything else is fixable.
Frequently Asked Questions
What’s the fastest way to check if my PC is healthy?
Download CrystalDiskInfo and check your drive health status. Then open HWiNFO and monitor temperatures under load for 20 minutes. If drives show “Good” status and temperatures stay under 85°C (CPU) or 90°C (GPU), your major components are healthy. This takes 30 minutes total.
Can I check PC health without downloading software?
Yes, but you’ll miss critical data. Windows includes CHKDSK for drives, Windows Memory Diagnostic for RAM, and SFC for system files. However, these tools don’t show SMART data, detailed temperatures, or component-level diagnostics. Built-in tools are better than nothing but incomplete for serious health checks.
How do I know if my computer is dying?
Dying computers show multiple symptoms: frequent crashes, blue screens, programs closing unexpectedly, slow performance despite low CPU usage, files corrupting, drives clicking or grinding, constant overheating, and failure to boot reliably. One symptom might be software. Multiple symptoms mean hardware failure. Run full diagnostics immediately and backup data.
Is my PC too old to check?
No. Older computers need more frequent health checks, not fewer. Components degrade with age. Check pre-2015 computers monthly instead of quarterly. Focus on drives (fail first), power supplies (capacitors age), and cooling (thermal paste dries). Even 10-year-old computers can run well with proper maintenance.
What causes most PC failures?
Storage drives cause 40% of failures, followed by power supplies (20%), RAM (15%), overheating damage (15%), and motherboard failures (10%). Regular health checks catch drive failures and overheating early. PSU and RAM failures often happen without warning, which is why testing twice yearly matters.
