Reliability Monitor is a built-in Windows 11 tool that turns your crash history into a day-by-day chart. To open it, press Win + R, type perfmon /rel and press Enter. Then click the first red X on the timeline and read what failed that day. This guide shows you how to read the chart, what the odd entries mean, and when running SFC and DISM is worth your time (and when it isn't).
TL;DR
- Open it with
perfmon /rel, or type "reliability" in Start search and pick "View reliability history." - The chart scores stability from 1 to 10 across the last 28 days. Find the day the line dropped and look at what happened then.
- Check Reliability Monitor first, then decide on repairs. SFC and DISM fix corrupted Windows files, not bad drivers or failing hardware.
- A recurring "Hardware Error" entry deserves attention, but it is often a GPU driver timeout rather than a dead component.
Table of Contents
- How to open Reliability Monitor in Windows 11
- What the chart is telling you
- A five-minute method for finding the cause
- When SFC and DISM are the right move
- Decoding "Hardware Error" entries
- Why Reliability Monitor is not working (or looks empty)
- Read it alongside other tools
- Our perspective: diagnose first, repair second
- Where Tempered fits
- FAQ
- Sources
How to open Reliability Monitor in Windows 11
There are two quick routes, and both are described by How-To Geek:
- Start search: type "reliability" and choose View reliability history.
- Run dialog: press Win + R, type
perfmon /reland press Enter.
The tool also sits in the old Control Panel under Security and Maintenance, but the Run command is faster and works the same on Windows 10. If you want to confirm the command is the documented launch method, the Windows Forum walkthrough also points to perfmon /rel.
The window opens straight onto a chart. No setup is needed, and nothing is installed. Windows has been collecting this data in the background all along.
What the chart is telling you
The top half of the window is a timeline. The bottom half is a list of events for whatever day you select.
The stability line
The blue line is the System Stability Index. It runs on a scale of 1 to 10, where 10 is the most stable, and the chart shows roughly the last 28 days. A healthy PC sits near the top with small dips. A PC with a driver problem often shows a staircase: each crash drags the line lower, and it recovers slowly.
Don't obsess over the exact number. The shape is what matters. You are looking for:
- a sudden drop on one day (something changed)
- a steady slide over a week (something keeps failing)
- a flat line at the bottom with the same icon repeating (one fault, over and over)
The icons
Four rows sit under the line, and each logs a different kind of event:
| Row | What it records | What to do with it |
|---|---|---|
| Application failures | Programs that crashed or stopped responding | Note the app name and the day. If it is one app, think update or reinstall. |
| Windows failures | Boot failures, unexpected shutdowns, blue screens | Check the day for a driver, update or power event. |
| Miscellaneous failures | Unexpected shutdowns and similar | Look for a pattern with sleep, heat or load. |
| Warnings | Failed updates, failed driver installs | Often the clue to what came right before a crash. |
Red X marks are failures. Yellow triangles are warnings. A blue "i" is informational, such as a successful install. The informational entries matter more than they look, because they show what changed just before things went wrong.
A five-minute method for finding the cause
The real value of the tool is correlation. As How-To Geek notes, experts use it to line up events, such as a driver update followed immediately by application crashes. You can do the same thing.
- Find the first drop. Click the earliest day where the line fell, not the worst day. The first failure is usually the closest to the cause.
- Read the day's list, top to bottom. Look at what was installed or updated just before the first failure.
- Click "View technical details" on a failure. The faulting module name (a
.dllor.sysfile) often names the guilty driver or program. - Count repeats. One crash in a month is noise. The same app or the same code six times is a signal.
- Pick the matching fix. One app: update or reinstall it. A driver: roll back or reinstall it. A damaged OS: now repair system files.
That last step is where most people start instead of finish. Which brings us to SFC and DISM.
When SFC and DISM are the right move
People searching for how to repair system files on Windows 10 or 11 often jump straight to the command line. Those tools do real work, but only on one kind of problem.
Microsoft describes System File Checker as a tool that scans for and repairs corrupted or missing system files. It does not fix a buggy graphics driver, a conflicting third-party program or a failing drive. If Reliability Monitor shows crashes tied to one game or one app, SFC will probably report nothing wrong and you will have spent half an hour for no gain.
Run the repair tools when the evidence points at Windows itself:
- Windows components crash or fail to start (Settings, Start menu, Explorer)
- Windows Update fails repeatedly
- Several unrelated built-in features misbehave
- Technical details name core Windows files
The sequence
Open Command Prompt as administrator (search "cmd," right-click, choose Run as administrator), then:
- Run
DISM /Online /Cleanup-Image /RestoreHealth. This repairs the Windows component store that SFC draws from. Windows Central explains the RestoreHealth option and how it uses different source files when the image has problems. - Run
sfc /scannow. Microsoft's SFC page documents this command. - Restart, then open Reliability Monitor again over the next few days and watch whether the failures stop.
People argue about whether SFC or DISM should go first, and you will find forum threads on both sides. Running DISM first is a common approach because SFC relies on the component store DISM repairs. Either way, both are safe to run, and neither touches your personal files.
The SFC and DISM commands can take a while, and progress bars sometimes seem stuck at a fixed percentage. Leave them alone unless nothing has moved for a very long time.
Decoding "Hardware Error" entries
This is the entry that worries people most, and it is one of the most searched questions about the tool. A Hardware Error in Reliability Monitor usually points to a failure caught by Windows error reporting, such as a disk, memory or GPU fault. When it recurs, it can mean the problem is beyond simple software corruption, and it may need hardware testing or a call to the manufacturer.
The name is scarier than the usual cause, though. Many of these are labeled LiveKernelEvent, with codes such as 141 or 1b8. In practice, a LiveKernelEvent often indicates a GPU driver timeout (TDR, Timeout Detection and Recovery) rather than a physically broken part. Windows hit a hang, recovered, and logged it. You can see this pattern in a 2026 Microsoft Q&A thread where a user reports black screens and signal loss while gaming, with the events logged as 1b8 and 141.
Use this to triage:
| Pattern | Likely direction |
|---|---|
| LiveKernelEvent during games or rendering only | GPU driver, overclock, temperature or power delivery |
| LiveKernelEvent at idle, with random restarts | Power supply, RAM or driver issue; test more widely |
| Kernel-Power 41 after hard shutdowns | The PC lost power or froze; look at PSU, heat, sleep settings |
| Disk-related errors alongside slow boots | Check drive health |
A reasonable order of attack for a recurring Hardware Error:
- Do a clean install of the current graphics driver. If you are comfortable with it, our guide to restarting the graphics driver covers the quick reset.
- Remove any overclock or undervolt, including the one-click "boost" profiles in some GPU utilities, and retest.
- Check the drive with our walkthrough on how to check SSD health.
- Test memory. Our piece on checking RAM usage in Windows helps you separate a real shortage from normal behavior.
- If errors continue after all that, treat it as hardware. A technician can test the power supply, RAM and board directly. If you want an idea of what a hands-on diagnostic visit looks like, this Swedish-language troubleshooting guide for a PC that won't start walks through common first checks before calling anyone.
If the machine restarts, shows a black screen or loses signal repeatedly, stop stress-testing it with games. Back up your files first, then diagnose.
Why Reliability Monitor is not working (or looks empty)
Sometimes the window opens with a blank chart or the message that not enough data is available. Common reasons:
- A fresh install or reset. The chart needs some days of history before it draws a meaningful line.
- Cleared logs. Clearing event logs or changing event policies can shorten the history you see. The view is a rolling window of about four weeks anyway.
- Services turned off. Reliability data depends on background services, so if you have disabled system services using a tweaking script, the log can stop collecting.
- A damaged system. If the tool itself will not open or errors out, this is one of the cases where running DISM and then SFC makes sense.
If the tool opens but you only see a few entries, that may simply be good news. A PC that rarely crashes has a boring log.
Read it alongside other tools
Reliability Monitor is a summary. It points at a day and a component, and then you go deeper. Event Viewer holds the raw detail behind each entry. Task Manager shows what is eating resources right now. Reliability Monitor tells you what happened over the past month. That split matters, because many slowdowns never crash anything and so never show up as red Xs.
For those, these pages help:
- Finding a CPU hog in Windows 11 when Task Manager shows a process pinned
- Checking available RAM rather than the percentage when memory looks high
- Diagnosing FPS drops with frametime traces when games stutter without crashing
- A repeatable VRAM stutter test for graphics memory limits
One point for gamers: LiveKernelEvents are not always fatal. They can be signs that Windows recovered from a hang. They tend to appear only under heavy load, so a PC that looks fine at the desktop can still be logging them after every long session.
Our perspective: diagnose first, repair second
Most PC "fixes" fail for the same reason: they begin with a remedy instead of a measurement. Someone runs a registry cleaner, a one-click optimizer, or a pile of command-line repairs, and the PC either stays the same or ends up in a new, harder-to-explain state.
Reliability Monitor is a good model of the opposite approach. It records what happened, with dates, and leaves the decision to you. That is how we think about Tempered, too. We look for the real constraint on a PC (a startup load, a background process, memory pressure, a driver issue), explain it in plain language, and propose a specific change you approve. Every change is tracked and can be undone with one click. We do not do registry cleaning, and we do not invent problems to look useful.
Our view on the SFC and DISM habit is simple. They are good tools used on the wrong problem too often. If the evidence points to a driver or an app, repairing Windows files is a detour. Check the timeline, name the suspect, then choose the smallest fix that tests your theory. If it does not work, you can reverse it and try the next one.
— Ian
Where Tempered fits
Reliability Monitor shows you the past 28 days. Tempered watches live CPU, RAM, GPU, disk and uptime, explains what is slowing the PC down, and suggests changes you can accept or skip, each with an undo. It also manages startup items and background processes, offers a Game Mode, and works without you bringing an AI API key.
The Free plan includes limited AI scans. Torque starts at $5.99 per month and Torque Max at $12.99 per month, both with higher scan limits and more features. You can see the plans or read more about the app before you decide.
FAQ
Is Reliability Monitor good for diagnosing problems?
Yes, for finding when and what, which is often the hardest part. It gives you a dated record of crashes, failed updates and driver events, so you can link a problem to a change. It will not always give you the root cause, so treat it as the first step before Event Viewer or hardware tests.
How do I get to Reliability Monitor in Windows 11?
Press Win + R, type perfmon /rel and press Enter. You can also type "reliability" into Start and pick "View reliability history." Both open the same window.
What does "Hardware Error" mean in Reliability Monitor?
It means Windows logged a fault that it linked to hardware, often a disk, memory or GPU. Many of these are LiveKernelEvents caused by a GPU driver timeout, not a failed part. If they keep recurring after a clean driver install and removing overclocks, get the hardware tested.
Should I run SFC and DISM every time Windows crashes?
No. They repair corrupted Windows system files only. If the failures trace back to one app or a driver, they will not help. Use them when Windows components themselves fail, running DISM's RestoreHealth and then sfc /scannow from an administrator Command Prompt.
Why is Reliability Monitor not working?
Usually there is little or no history yet, the logs were cleared, or the related services were disabled. If the window will not open at all, repair system files with DISM and SFC, then try again.
Sources
- How to View Reliability History in Windows (NinjaOne)
- Windows Reliability Monitor guide (How-To Geek)
- How to Use Windows Reliability Monitor to Find Stability Drops (Windows Forum)
- Using System File Checker in Windows (Microsoft Support)
- Use the System File Checker tool (Microsoft Support)
- Use the DISM command tool to repair a Windows 10 image (Windows Central)
- Why does my PC randomly restart? Live Kernel Event (Microsoft Q&A)
- Recurring LiveKernelEvent 1b8 hardware error (Microsoft Q&A)