Short answer: Click Cancel and do not write anything else to the disk. The message „You need to format the disk before you can use it" and a RAW status do not mean your data is gone—Windows simply cannot find a file system structure it can work with. The safe approach is to create a sector-by-sector copy on another storage device and recover data from that copy, rather than formatting the disk, running CHKDSK, or scanning the original.
This article is not a guide to getting the disk working again at any cost. It is a decision-making guide: what to do in the first minute, what can irreversibly overwrite your data, and when to stop trying and have the device diagnosed. There is a huge difference between „Windows cannot see the file system„ and „the data is permanently lost"—and a single click can turn the former into the latter. If the data is not backed up, the safer option is professional data recovery in a laboratory; for portable and desktop USB drives specifically, see external hard drive data recovery.
Disk asks to be formatted: what to do immediately
Windows offers to format a drive when it can see the partition but cannot mount it. This is fundamentally different from a disk that does not appear at all. The message alone does not prove that the data has disappeared; only the structure the system uses to locate it may be damaged.
First steps:
- Cancel every prompt to format or initialize the disk.
- Do not save anything to the disk—not even „just temporarily."
- Do not run
chkdsk, volume repair, or a partition recovery tool. - Do not reconnect the disk repeatedly in the hope that it will work next time.
- If it clicks, scrapes, repeatedly disconnects, or visibly struggles to read, switch it off.
- Make a note of what happened before the failure: a fall, power outage, unplugging without safe removal, an update, partition work, or removal from a NAS.
The last point deserves just as much attention as the others. „The disk dropped out of a Synology array„ and „the laptop fell off a table" lead to two entirely different procedures. Without that information, the first hour of diagnosis may be wasted working in the dark.
What RAW means
RAW is not a file system like NTFS, exFAT, or FAT32. It is Windows’ way of saying, „I cannot see a structure on this partition that I know how to read." Nothing more and nothing less—it says nothing about the contents of the platters or NAND chips.
Only part of the metadata may be damaged. NTFS stores key information in the volume boot sector and the MFT file table, while the partition layout is stored in the partition table (MBR or GPT). If the boot sector or an important part of the MFT is damaged, Windows may lose its entry point and report the partition as RAW even though some file contents remain readable on the device. NTFS also uses a backup boot sector and the $MFTMirr file, but their presence alone does not guarantee complete recovery of file names and directories.
The problem is that the same message can have very different causes—and each one tolerates a different amount of further use:
| What happened | Typical symptoms | Risk of switching it on again |
|---|---|---|
| Damaged file system | The disk reports the correct model and capacity but appears as RAW | Any write may overwrite metadata needed for recovery |
| Bad HDD sectors | Reading slows down, the disk disconnects, or the system freezes | The damage may spread, and continued reading may further damage the surface |
| Faulty USB cable or enclosure | The disk works intermittently and then disappears | Lower, provided the disk makes no unusual sounds |
| Read/write head failure | Clicking, failure to spin up, or repeated spin-up attempts | High—the heads may scratch the platters |
| SSD or controller failure | The disk disappears, reports the wrong capacity, or appears as RAW | Its condition may change without warning, and the disk may „die" between power cycles |
| Disk removed from a NAS | Windows offers to format it even though the disk may be healthy | Formatting will damage a RAID or SHR member |
The final row is the most deceptive. Windows treats a perfectly functional disk from a Synology NAS in exactly the same way as a failing HDD—and the message gives the user no way to tell the difference.
Why you should not click Format
Formatting is not a repair tool. It is a write operation. It creates a new, empty file system structure exactly where the original one was stored—overwriting the records that could otherwise be used to locate file names, folders, and their positions.
Even a quick format has consequences. It does not overwrite the entire volume sector by sector, but it creates new file system structures and removes the original file table and root directory. Depending on the original and new file systems, this may overwrite metadata needed to recover file names, folders, and file locations. Recovery may then be reduced to searching for files by signature, producing names such as file0001.jpg with no original directory tree.
On SSDs, recovery may also be severely limited by the TRIM command. The operating system uses it to tell the storage device which areas are no longer needed, allowing the device to release them. Whether and when this happens depends on the operating system, controller, firmware, and specific operation, so no recovery window can be promised after a quick format. The safe response is to stop using the disk immediately.
If you have already clicked Format, it is not necessarily the end. Stop using the disk immediately, however—every installation, copy operation, and recovery attempt run directly on that device reduces what may still be recoverable.
When to check the cable—and when not to power the disk on again
It is worth making one controlled test—using a different port, a known-good data cable, or another computer—only if the disk:
- spins up normally and remains connected,
- makes no clicking or scraping sounds,
- does not disconnect while being read,
- does not cause the system to freeze,
- has not been dropped or exposed to liquid.
Connect it directly to the computer, not through a monitor, dock, or passive hub. With USB-C, make sure the cable actually carries data—some cables in your drawer may support charging only. We explain why this connector can cause such problems in why we dislike USB-C.
If replacing the cable makes no difference, further testing at home will provide no new information—it will only put additional strain on the device. With an HDD that has bad sectors, that strain is not harmless: every attempt to read a damaged area sends the heads over the problematic track again.
If the disk clicks, repeatedly spins up, or sounds different after being dropped, switch it off and do not try anything else. The procedure for this situation is described in what to do when a hard drive clicks. If the disk does not appear at all, continue with external drive not loading or not detected.
When not to try it yourself—and what we will do
With a disk reported as RAW, a DIY recovery attempt is particularly dangerous when:
- the disk clicks, scrapes, or repeatedly spins up—RAW is only a symptom of faulty heads, and further reading damages the platters,
- the disk is detected intermittently and then disappears or reports the wrong capacity—the fault is in the hardware, not the file system,
- the disk was dropped, suffered a power surge, or has already been subjected to CHKDSK or formatting—every additional write or scan reduces what can still be recovered.
In these cases, switch the disk off and bring or send it in for external hard drive data recovery, or see HDD data recovery for internal hard drives. The process has three steps:
- Intake—bring the disk to our laboratory in Brno-Židenice or send it to us; we record what happened before the failure and what you have already tried.
- Diagnosis—we determine whether the problem is structural or hardware-related and what can be recovered. Initial diagnosis of a standard HDD or SSD may be free; unusual cases—such as previous formatting or overwriting, a disk opened by another repair service, or complex analysis—cost 1 210 Kč including VAT (1 000 Kč excluding VAT), and you will be informed in advance.
- Price quote, then recovery—you receive the recovery price before work begins, and we do nothing further to the disk without your approval. Indicative prices by fault type are listed in our data recovery price list.
Is it urgent? Call our 24/7 line at +420 775 556 063.
Why not to run CHKDSK, TestDisk, or recovery software on the original disk
chkdsk is a file system consistency repair tool, not a data recovery tool. On a volume reported as RAW, it will usually stop with a message saying that „CHKDSK is not available for RAW drives" and do nothing. If it does start—typically on a volume that is still recognized as NTFS but is damaged—it writes to the disk: it repairs structures, releases records it considers invalid, and stores recovered fragments in the found.000 folder as file0000.chk files. From a recovery perspective, this is the worst possible outcome: the data is still on the disk, but fragmented and stripped of its original names.
The same principle applies to macOS First Aid and Linux fsck. They can help with logical inconsistencies on a healthy device. On a disk with bad sectors, they add load and write operations exactly where they are least desirable.
TestDisk is a good tool, and analysis alone does not necessarily write anything. It only writes changes to the partition table or boot sector after the Write option is explicitly confirmed, but an incorrect write can complicate subsequent recovery. Recuva, Disk Drill, R-Studio, and DMDE are useful for purely logical data loss on healthy media. None of them should scan a physically failing disk directly, however: a deep scan means hours of reading the entire surface, which is exactly what an unstable disk may not survive.
The safe procedure therefore has two separate stages:
- First, create a sector-by-sector copy (image) on another storage device.
- Only then search the copy for partitions and files—using any tool, repeatedly, without putting the original at risk.
Unstable media should be read with a tool that can skip damaged areas, return to them later, read from the opposite direction, and maintain a map of what has already been copied. A typical example is GNU ddrescue—and that map is crucial because it allows the process to be interrupted and resumed without making the disk struggle again with areas where it has already failed. Even the best software cannot repair a disk with faulty heads; that requires work in a clean environment, not another program. Bad sectors on an HDD explains what can and cannot be done with a damaged surface.
Disk removed from Synology or QNAP: a warning that does not indicate failure
This is the most common false alarm. Synology DSM and QNAP QTS build their volumes on Linux software RAID (md), with LVM and an ext4 or btrfs file system layered on top. Windows cannot read any of these layers natively. When you connect such a disk to a computer, the operating system sees a partition it does not understand and offers to format it. The disk itself may be perfectly healthy.
The brand on the label tells you nothing about this. WD Red, Seagate IronWolf, and Toshiba N300 drives all behave the same way in Windows because the data format—not the manufacturer—is what matters. We compare their suitability for NAS use in WD Red vs. Seagate IronWolf vs. Toshiba N300.
With an array, recovering one disk in isolation is not enough. Reconstruction requires the order of the members, the array geometry, and the condition of every device. With Synology SHR, the data is not necessarily stored in a single contiguous RAID but may be divided among several RAID groups above LVM, so one removed disk cannot provide a complete set of files even in theory. Do not click Initialize or options such as „Recover," do not reinstall the disks in a random order, and above all, do not start a rebuild on an array that is already reporting a problem—we explain why in why rebuilding can destroy a RAID array. For specific conditions, see Synology: storage pool degraded and QNAP: RAID group degraded.
What professional RAW disk recovery looks like
The first step is to determine the type of failure: a logical error, a communication problem, surface degradation, or a mechanical fault. At a minimum, the diagnosis should answer the following questions:
- Does the disk report the correct model and capacity?
- Can it be read reliably, or does it disconnect during reading?
- Is the number of unreadable sectors increasing or stable?
- Is only the partition structure damaged, or is the hardware itself failing?
- Is the volume encrypted with BitLocker?
- Which data is the highest priority for the business?
The BitLocker question is not a formality. An encrypted volume without its key looks like random data when read, and without the recovery key, nobody can extract anything from it—even if the platters are healthy. BitLocker explains where to look for the key.
If the disk is readable, an image is created and the file system is reconstructed from that image. A physical failure may require work on the electronics or heads. Recovered data is always saved to another device—never back to the original disk.
It is worth setting priorities in advance. Accounting databases, active project documentation, and current client work generally matter more than an archive of installation files. A priority list is especially important when the entire disk cannot be read—which is common when bad sectors are involved.
How much it costs and what to check beforehand
The type of failure determines the price, not the capacity. A logically damaged disk that can be read reliably is very different from an HDD with faulty heads or extensive surface damage. With SSDs, the controller type, encryption, and the availability of a method for reading the memory outside the controller all play a role.
Before placing an order, clarify:
- whether diagnosis is charged and exactly what it includes,
- whether you will receive a price or price range before recovery begins,
- what counts as a successful recovery,
- whether you will receive a list of recovered files for review,
- which storage device will be used to return the data,
- how the provider handles confidential business data,
- what will happen to the original disk and working copies after delivery.
No reputable provider can give a reliable price based only on a RAW warning, because at that point nobody knows whether the problem is a single damaged structure or the drive heads. How much disk data recovery costs explains what actually determines the price; the disk data recovery service page describes the process step by step.
You can recognize a sensible approach by the questions being asked, not by the promises being made. The provider should ask about the symptoms, previous interventions, encryption, and the importance of the data. They should separate work on the original device from file recovery and should not ask you to approve unlimited costs before diagnosis. Warning signs include a vague price with no explanation, a request for the password to a general business account instead of the specific BitLocker recovery key—and, above all, a proposal to „format it first and see what happens."
What to prepare before handing over the disk
Write down how the problem appeared and everything that has already been done to the disk—including unsuccessful attempts. Concealing the fact that CHKDSK was run does not save time; it only sends the diagnosis down a blind alley.
For an external disk with its own power supply, include the original adapter. For an encrypted device, locate the BitLocker recovery key, but do not send it by unencrypted email. For a NAS, label the disks by their bay positions and do not change their order—and do not hand over only the disk that the system marked as faulty. Reconstructing the array generally requires the complete set.
Frequently asked questions
Should I click Format when Windows says the disk cannot be used otherwise?
No, not if you need the data stored on it. Formatting writes a new file system over the structures used to locate the original file names and folders. The warning also says nothing about the disk’s health—a perfectly functional disk removed from a NAS can trigger it too.
Can data be recovered from a disk that appears as RAW?
Often, yes, but it depends on the cause. If the file system is damaged but the storage device itself is healthy, most of the data is usually readable. Bad sectors, head failure, SSD failure, or previous writes to the disk reduce the chances of recovery.
Can CHKDSK repair a RAW disk?
It is not a data recovery tool. It usually will not run at all on a volume reported as RAW, and when it does run, it writes to the disk—it may place recovered fragments in the found.000 folder as files without their original names. Working from a sector-by-sector copy is always safer than repairing the original.
What if I have already quick-formatted the disk?
Stop using it and do not save anything to it. On an HDD, some data may remain recoverable, but the original file names and folder structure are often damaged. On an SSD, TRIM may severely limit recovery—released blocks are cleared and return zeros when read.
Why does Windows offer to format a disk from Synology or QNAP?
Because Windows cannot read Linux partitions, ext4, btrfs, or the software RAID and LVM layers used to build NAS volumes. The format prompt therefore does not necessarily indicate any fault—but confirming it will damage a member of the original array.
Verified sources
- Microsoft Learn: Format command and quick-format behavior
- Microsoft Learn: TRIM/Unmap notifications
- GNU ddrescue Manual
- Official TestDisk documentation: file system repair
Where to find us
ITHOPE is a company based in Brno, Czechia: we do data recovery from disks, RAID and NAS, PC and laptop service and IT for business, all in our own lab at Viniční 3587/216, 615 00 Brno.
Phone +420 775 556 063 (24/7 line). You can bring the device in person, or send it by courier (Zásilkovna, PPL).