Windows Search Not Finding Files
Fix the index and the Windows Search service, stop the indexer eating your CPU, and search inside your files with a tool that doesn't depend on the Windows index. It runs locally on your PC — nothing gets uploaded.
The short answer
- ✓ First, fix Windows Search: run the Search and Indexing troubleshooter, restart the Windows Search service, then rebuild via Control Panel > Indexing Options > Advanced > Rebuild. Make sure the folder is in your indexed locations and that 'Index Properties and File Contents' is enabled so search looks inside files.
- ✓ Most misses come from a corrupt or incomplete index, the folder not being in the indexed locations, file-contents indexing being turned off, the Windows Search service being stopped, or the index hitting its item limit. A bad index can also make SearchIndexer.exe hammer your CPU and disk.
- ✓ If it keeps failing — or the indexer keeps eating CPU — the durable fix is a search that doesn't use the Windows index at all. A live, ripgrep-style search that looks inside file contents (text, code, and documents like PDF and DOCX), with no always-on indexer to corrupt or to drain your CPU. FileHop does this locally, with no uploads, on Windows and Mac.
Quick fix checklist
Run through these in order. The first few resolve most cases; the rest catch the common edge cases.
Run the Search and Indexing troubleshooter (Settings > System > Troubleshoot > Other troubleshooters > Search and Indexing).
Restart the Windows Search service: press Win+R, type services.msc, find 'Windows Search', set it to Automatic and restart it.
Rebuild the index (steps below) — Control Panel > Indexing Options > Advanced > Rebuild.
Confirm the folder is in your indexed locations (Indexing Options > Modify) and add it if it's missing.
Enable content search: Indexing Options > Advanced > File Types tab > select the type > 'Index Properties and File Contents', and in Folder Options > Search turn on 'Always search file names and contents'.
Free up space if the drive is nearly full, then let indexing finish before testing again.
Why Windows Search misses files that clearly exist
Windows Search doesn't read your files on the spot — it answers from a background index that SearchIndexer.exe builds over time. When that index is corrupt, incomplete, or doesn't cover the folder, search returns nothing even though the file is right there. These are the usual causes:
Corrupt or incomplete index
The index can get corrupted or fall out of sync after a Windows update, an interrupted build, or a crash. Search keeps answering from the broken index, so files that exist never show up. Rebuilding the index is the standard fix.
The folder isn't in your indexed locations
Windows only indexes the locations listed in Indexing Options. If your files live somewhere outside those locations (a different drive, a custom folder), search either skips them or falls back to a slow, incomplete non-indexed scan.
File-contents indexing is turned off
By default many file types are indexed by properties only, not contents. If 'Index Properties and File Contents' isn't enabled for the file type, searching for a phrase inside the document returns nothing even when the index is healthy.
The Windows Search service is stopped or disabled
If the Windows Search (WSearch) service isn't running, the index can't update and search behaves erratically or returns nothing. A disabled service, often after troubleshooting or a third-party 'optimizer', is a common culprit.
The index hit its item limit or ran out of room
Very large libraries can exceed practical index limits, and a nearly full system drive starves the index of working space. Either way, indexing stalls or skips files and results go missing.
Some folders are excluded, or it works in File Explorer but not Start
Folders can be excluded from indexing deliberately or by accident. Start-menu/taskbar search and File Explorer search use the index differently, so you can get results in one place and nothing in another for the same file.
Network drives and cloud-only files
Mapped network drives often aren't indexed, and OneDrive 'Files On-Demand' items that are online-only are just placeholders on disk, so their contents can't be searched until they're downloaded locally.
Worth knowing: Microsoft itself publishes 'Fix problems in Windows Search' and 'Troubleshoot Windows Search performance' docs, and there are active Microsoft Q&A and forum threads about search 'not finding files even though I can see them' after recent Windows 11 updates. If a rebuild doesn't stick, you're not imagining it — and a search that doesn't rely on the index sidesteps the whole problem.
Why SearchIndexer.exe uses high CPU or disk — and what to do
The other half of this problem is the flip side of the same index. SearchIndexer.exe is the process that builds and maintains the Windows Search index, and when it goes wrong it can sit at high CPU or hammer your disk for long stretches.
It's re-indexing a corrupt index
After a rebuild or a crash, SearchIndexer.exe re-scans everything from scratch — which is exactly the heavy CPU/disk activity people notice. Normally it settles once the build finishes, but a repeatedly corrupting index can leave it churning indefinitely.
Too many or too-large indexed locations
Pointing the index at huge media libraries, archives, PST mail files, or whole drives full of large documents gives it far more to chew through, keeping CPU and disk busy. Trimming indexed locations to just what you actually search reduces the load.
SSD wear and constant disk activity
On SSDs, people worry about the constant writes from re-indexing. Limiting indexed locations and stopping the cycle of corrupt-then-rebuild reduces the churn. To fix it: rebuild once cleanly, run the troubleshooter, and pare back what gets indexed.
But there's a more direct way out of both pains at once: a search that has no always-on indexer doesn't have a SearchIndexer.exe to spike your CPU in the first place, and because it doesn't depend on the index, there's nothing to corrupt and rebuild. That's the durable fix below.
How to rebuild the Windows Search index
This is the canonical fix from Microsoft's own guidance. It throws away the existing index and builds a fresh one.
Open Control Panel and go to Indexing Options (search for 'Indexing Options' if you don't see it).
Click Modify and make sure the folders and drives you want to search are checked, then click OK.
Click Advanced, then on the Index Settings tab click Rebuild and confirm. Windows discards the old index and starts a fresh one.
Still in Advanced, open the File Types tab, select the file types you care about, and choose 'Index Properties and File Contents' so search looks inside the files, then click OK.
Leave the PC on while indexing finishes (it can take a while and you'll see SearchIndexer.exe busy during this time), then test your search again.
For power users (PowerShell)
Prefer the command line? You can search text inside files without the index using Select-String -Path C:\path\*.txt -Pattern "some phrase", or reset Windows Search with the built-in ResetWindowsSearchBox.ps1 script. These help, but Select-String is slow over large trees and the reset still relies on the very index that's failing — if a rebuild doesn't stick, the durable answer below is a search that doesn't use the index at all.
The durable fix: a search that doesn't rely on the Windows index
Every fix above sends you back to rebuilding the same index that just failed — or to the indexer that's eating your CPU. If you keep hitting the wall — or you specifically need to find the file that contains a phrase — the more reliable approach is a search that doesn't depend on the Windows index at all.
FileHop searches a folder you point it at, live. There's no background SearchIndexer.exe building a system-wide index that can go stale or corrupt, and nothing to rebuild. Because it reads the files at search time, the results are never out of date — what's on disk is what you get — and there's no always-on indexer spiking your CPU or wearing your SSD when the PC is idle.
Crucially, it searches inside file contents, not just filenames. That covers plain text, code, Markdown, JSON, CSV, and logs — and it also reads inside documents by extracting their text: PDF, DOCX, PPTX, XLSX, XLS, and ODS. So you can find the one contract or note that mentions a phrase, even when you've forgotten its filename. It runs locally on your PC (and Mac), with no uploads and no account.
Where this fits — honestly
FileHop is not a drop-in replacement for the Start-menu or taskbar search box. You point it at the folder or drive where your files live, rather than searching the whole OS automatically. A few real limits to know: scanned, image-only PDFs with no text layer return nothing (they'd need OCR, which the content search doesn't do); legacy binary .doc and .ppt aren't readable (modern .docx/.pptx/.xlsx are); and there's a 10MB default per-file size cap you can raise. Within those limits, it's the most reliable way to search inside your files when the Windows index lets you down.
How searching inside files with FileHop works
Three steps, all on your PC.
Open the folder where your files live
Point FileHop at the folder or drive you want to search — your Documents, a project folder, an external drive, wherever the file is.
Type a filename or a phrase inside the file
Search by filename with glob patterns, or by a phrase, word, or regex that's inside the file. Toggle whole-word or case-sensitive matching as needed.
Jump straight to the match
Results show the file path, the line number, and the surrounding context — so you land exactly on the match. Nothing was uploaded; it all happened locally.
Windows Search vs. the alternatives — an honest comparison
Several tools claim to fix Windows search. The catch most reviews skip: the go-to fast search (Everything) is filename-only fast, and its own docs call content search slow. Here's how the real options line up on the axes that actually matter.
| What matters | Windows Search | Everything | Agent Ransack | DocFetcher | FileHop |
|---|---|---|---|---|---|
| Depends on the Windows index? | Yes (it is the index) | No (its own filename index) | No (live scan) | No (builds its own index) | No |
| Always-on indexer using CPU when idle? | Yes (SearchIndexer.exe) | No (filename index only) | No | No (manual index) | No always-on indexer |
| Fast filename search? | Only if indexed | Yes (instant) | Yes (live) | Via its index | Yes (per-folder index) |
| Fast content (inside-file) search? | Only if indexed; non-indexed is slow | Slow (its docs say so; not indexed) | Yes (live) | Yes (via its index) | Yes (live, ripgrep-style) |
| Searches inside PDF / DOCX text? | Only with the right IFilter | No | Limited | Yes | Yes |
| Local / no uploads? | Yes | Yes | Yes | Yes | Yes |
| Price | Built in | Free | Free (Pro paid) | Free (open source) | Free |
The honest read: Everything is the beloved instant filename search on Windows — but its own documentation says content search ('content:') is extremely slow and that file contents are not indexed, so it's filename-only fast. Agent Ransack does live content search with a GUI and is genuinely useful and free (the Pro 'FileLocator' is paid). DocFetcher is free and open source and reads inside documents, but you build and maintain its own index first. Windows Search is index-dependent and brings the SearchIndexer.exe CPU cost. FileHop's niche is the gap they leave: fast live content search, inside documents too, with no Windows-index dependency and no always-on indexer.
Why this approach holds up
No dependence on the Windows index — nothing to corrupt or rebuild.
No always-on indexer — no SearchIndexer.exe draining CPU or disk when the PC is idle.
Searches inside file contents, not just names.
Reads inside PDF, DOCX, PPTX, and XLSX text too.
Live results — what's on disk is what you find, never stale.
Stays on your PC — no uploads, no account. Free, for Windows and Mac.
Still not finding files? Troubleshooting the stubborn cases
If a rebuild didn't solve it, these are the edge cases that trip people up.
Search works in some folders but not others
The folders that work are in your indexed locations; the ones that don't aren't, or are excluded. Add the missing folder in Indexing Options > Modify, or search it directly with a tool that doesn't use the index, like FileHop, which scans the folder live.
Search stopped finding files after a Windows update
Updates can corrupt or reset the index. Run the Search and Indexing troubleshooter, restart the Windows Search service, and rebuild the index. If it keeps breaking after updates, an index-independent search sidesteps the cycle entirely.
The index is stuck or shows 0 items indexed
Confirm the Windows Search service is running and set to Automatic, free up disk space, then rebuild. If the count stays at zero, the index is failing to build — search the folder directly with a live tool while you sort it out.
Content search returns nothing for a PDF or document
Either the file type isn't set to 'Index Properties and File Contents', or it's a scanned, image-only PDF with no text layer (which needs OCR first), or it's a legacy .doc/.ppt that can't be read. Enable content indexing for the type, run OCR on scanned PDFs, and use modern .docx/.pptx/.xlsx where possible.
Frequently Asked Questions
Why is Windows Search not finding files that I can see?
Windows Search answers from a background index built by SearchIndexer.exe rather than reading your disk live, so when that index is corrupt or incomplete, the file won't show up even though it's there. Other common causes are the folder not being in your indexed locations, file-contents indexing being turned off, the Windows Search service being stopped, or the index hitting its item limit. Run the Search and Indexing troubleshooter, restart the service, and rebuild the index; if it keeps failing, a search that doesn't use the Windows index (like FileHop) will still find the file.
How do I rebuild the Windows search index?
Open Control Panel > Indexing Options, click Advanced, and on the Index Settings tab click Rebuild and confirm. Windows discards the old index and builds a fresh one — leave the PC on while it finishes, since SearchIndexer.exe will be busy during that time. While you're there, click Modify to confirm your folders are included, and use the File Types tab to enable 'Index Properties and File Contents' so search looks inside files.
How do I make Windows search inside file contents?
Open Indexing Options > Advanced > File Types, select the file type, choose 'Index Properties and File Contents', and click OK; then in File Explorer's Folder Options > Search, turn on 'Always search file names and contents'. Note that content search over non-indexed folders is slow, and Everything's content search is slow too. For fast, reliable content search, FileHop reads inside text, code, PDF, DOCX, PPTX, and XLSX live, without depending on the Windows index.
Why does File Explorer search work in some folders but not others?
The folders that return results are in your indexed locations; the ones that don't are either not indexed or explicitly excluded. Searching a non-indexed folder falls back to a slow, incomplete scan that often misses things. Add the folder in Indexing Options > Modify, or point a tool that searches live (like FileHop) directly at the folder so it doesn't matter whether the index covers it.
Why is SearchIndexer.exe using so much CPU or disk?
SearchIndexer.exe is the process that builds and maintains the Windows Search index. High CPU and disk usually mean it's re-indexing after a corrupt index, an update, or a crash, or that you're indexing too many or too-large locations (media libraries, archives, PST files). It normally settles once the build finishes; to reduce it, rebuild once cleanly, run the troubleshooter, and trim indexed locations to what you actually search. A search with no always-on indexer (like FileHop) avoids this CPU and SSD-wear cost entirely.
How do I search text inside PDFs and Word documents on Windows?
Windows Search can search inside documents only if the right IFilter is installed and 'Index Properties and File Contents' is enabled, and even then non-indexed folders are slow. FileHop extracts the text from PDF, DOCX, PPTX, XLSX, XLS, and ODS files and searches inside it live, showing the file path, line number, and surrounding context. The one limit: a scanned, image-only PDF has no text layer, so it needs OCR first before any tool can read words inside it.
Is there a Windows search that doesn't use the index?
Yes. Agent Ransack and DocFetcher both search without depending on the Windows index (DocFetcher builds its own index you maintain). FileHop searches a folder you choose live, with no background index to corrupt or rebuild and no always-on indexer, so results are never stale and there's no SearchIndexer.exe CPU hit. Everything also has its own filename index independent of Windows Search, but its content search is slow and not indexed.
Why does Everything not search file contents quickly?
Everything is built around an instant filename index, and its own documentation states that file contents are not indexed and that content search ('content:') is extremely slow because it reads each matching file at search time without a content index. That makes Everything excellent for finding files by name but a poor fit for finding a phrase inside many documents. A ripgrep-style live content search like FileHop is built for exactly that case, including inside PDF and DOCX text.
How do I search files on a network drive?
Mapped network drives often aren't indexed by Windows Search, so files there go missing from results, and indexing a network share is frequently not possible at all. The reliable approach is to point an index-independent search directly at the mounted drive. FileHop can open a network or external drive and search inside the files there live, without needing Windows to index the volume first.
Does searching my files upload them anywhere?
With Windows Search and with FileHop, no — both run entirely on your PC and don't upload your files. Online search or conversion services do upload, which is a privacy concern for personal documents. FileHop keeps everything local: no account, no upload, and no cloud processing, on both Windows and Mac.
Find the file Windows Search can't
FileHop searches inside your files — names and contents, including PDF and DOCX — live and locally, with no dependence on the Windows index and no always-on indexer. Free, no uploads.
Download FileHop Free - Windows & Mac