A monitor that randomly goes black — screen cuts to black for 1–5 seconds then returns, or cuts to black and doesn’t come back until the cable is wiggled — is deeply annoying and has a specific diagnostic that tells you which layer to fix. We go deeper on the whole subject in our Complete Guide to Fixing Windows, Browser, and Software Errors.
The first thing to establish: does the black screen affect only one monitor in a multi-monitor setup, or all monitors simultaneously? If only one monitor: it’s either that monitor specifically, the cable to it, or the GPU output port it’s connected to. If all monitors black out simultaneously: the GPU, its driver, or the display driver is resetting.
The second thing: does the screen go completely black (no backlight), or does the backlight stay on but showing a black image? Backlight on = signal lost (cable, GPU). Backlight off = display going into sleep or power off mode (usually power management or a failing display panel).
Cable and connection — check first
A partially seated connector, a damaged cable, or a cable that’s too long for the resolution and refresh rate combination is responsible for a large fraction of random blackout events. Press both ends of the cable firmly into their ports — DisplayPort connectors have a latch that clicks; HDMI doesn’t but can be loose. Try a different cable entirely. Try a different port on the GPU.
For DisplayPort specifically: try switching to HDMI, or vice versa. Some GPU outputs have intermittent issues that don’t show in testing but fail randomly during use. Using a different port and a different cable simultaneously rules out both variables at once.
GPU driver — the most common software cause
Random black screens that are caused by the driver resetting (TDR — Timeout Detection and Recovery) happen when the GPU driver encounters an error, fails to respond to Windows within 2 seconds, and Windows forces a driver reset. The screen goes black briefly while the driver restarts, then returns with a “Display driver stopped responding and has recovered” notification in the corner.
Update the GPU driver from the manufacturer’s website (NVIDIA, AMD, or Intel — not Windows Update for this). If the problem started after a driver update: roll back in Device Manager → Display Adapters → right-click GPU → Properties → Driver → Roll Back Driver. Clean driver reinstallation with DDU (Display Driver Uninstaller) eliminates driver remnants from previous versions that can cause TDR events.
Adaptive sync (G-Sync/FreeSync) and random blanks
Adaptive sync implementations that aren’t fully compatible with the monitor — G-Sync on a FreeSync-only monitor, FreeSync outside the supported VRR range, or adaptive sync over HDMI instead of DisplayPort — cause brief blank events when frame rate transitions through the sync boundary. The screen goes black for a fraction of a second at specific frame rates.
Test: disable G-Sync or FreeSync in the GPU control panel → watch for blank events. If they stop: the adaptive sync configuration was causing them. Correct configuration: use DisplayPort instead of HDMI for adaptive sync, confirm the monitor model is on the NVIDIA G-Sync Compatible or AMD FreeSync certified list, and configure the frame rate cap to stay within the monitor’s VRR range.
Monitor power settings and auto-off
Monitors have their own internal timer to go into standby. If this timer is set very short (5–10 minutes) and there’s a period without signal variation (a static image, a pause), the monitor may self-sleep while Windows thinks it’s still displaying. The result: a black screen that comes back when you move the mouse.
Access the monitor’s OSD (on-screen display) via its physical buttons → find Power or Display settings → increase or disable the monitor’s internal sleep timer. This is separate from Windows’ screen timeout — Windows can be set to never turn off the display while the monitor’s own timer is set to 10 minutes.
GPU hardware instability
Overheating GPUs cause display resets that look like driver TDR events but don’t produce the “driver recovered” notification. The GPU throttles, loses the display signal briefly, then recovers. Monitor GPU temperature during a session using MSI Afterburner or HWiNFO64 — if temperatures regularly exceed 85°C and black screens correlate with temperature spikes: thermal throttling is the mechanism.
Cleaning GPU fans and heatsink (compressed air through the vents, or GPU disassembly for a proper clean on older cards) often dramatically reduces temperatures. Reapplying thermal paste on a GPU that’s 4+ years old can drop temperatures by 15–20°C. For desktop GPUs: ensure adequate case airflow and that the GPU isn’t starved of cool air by poor cable routing.
HDR and high refresh rate transitions
Switching between HDR and SDR content, or between high refresh rate and standard content, causes the monitor to renegotiate its display parameters. During this renegotiation — which takes 1–3 seconds — the screen goes black. On some monitor and GPU combinations, this renegotiation happens more frequently than expected, producing repeated brief blackouts that aren’t actually crashes.
Settings → System → Display → Windows HD Color → if HDR is enabled and blackouts correlate with switching applications: disable HDR and test for stability. For high refresh rate monitors: set a fixed refresh rate in Settings → Display → Advanced display rather than letting Windows vary it — constant refresh rate eliminates the renegotiation triggers.
HDCP and content protection resets
HDCP (High-bandwidth Digital Content Protection) handshakes happen when protected content starts playing. If the HDCP handshake fails (monitor or cable doesn’t fully support HDCP, or there’s a compatibility issue), the display connection resets, causing a brief black screen. This happens specifically when starting protected video (Netflix, Prime Video, encrypted blu-ray) rather than randomly throughout use.
If black screens correlate with starting protected content: the HDCP implementation on the cable, monitor, or GPU port has an issue. Try a certified HDMI 2.0 or HDCP 2.2 compatible cable rather than a generic one.
Our guide on screen flickering covers the Task Manager diagnostic that distinguishes driver-level flickering from application-level flickering — the same diagnostic applies to black screen events that may be partial flickers rather than full signal loss. For the GPU temperature and thermal management that causes display issues, our GPU performance guide covers the monitoring tools and cooling approaches. Microsoft’s TDR documentation covers the TdrDelay and TdrDdiDelay registry values that control how long Windows waits before forcing a driver reset — adjusting these can reduce false TDR events on marginal hardware.
Event Viewer for TDR events
Event Viewer → Windows Logs → System → filter for Source “Display” or “nvlddmkm” (NVIDIA) or “atikmdag” (AMD). TDR events appear as errors with Event ID 4101 (NVIDIA) or similar. These events include the timestamp (confirming correlation with the black screen events), the recovery status, and sometimes the specific GPU operation that triggered the reset.
If TDR events appear: GPU driver issue confirmed. If no TDR events appear at the time of the black screen: the cause is not a driver crash — it’s a signal loss (cable, port, or monitor-side issue) or a Windows screen timeout. This distinction is critical for directing troubleshooting correctly.
PSU insufficiency on desktops
Desktop machines with power supplies that are borderline for the installed hardware can produce display dropouts under load. When the GPU demands peak power (during demanding game scenes, large model loads, or simultaneous GPU encoding and rendering), the PSU can’t maintain voltage and the GPU briefly loses power stability, causing the display output to drop and reconnect.
Signs: black screens that correlate specifically with heavy GPU workloads, system stability is fine during light use, and the display always comes back. Power draw monitoring (HWiNFO64 shows GPU power consumption in Watts) during the black screen event — if the GPU was drawing near or above the PSU’s GPU budget at the moment of the black screen, PSU capacity is the issue. A higher-wattage PSU with better transient response is the fix.
USB-C and Thunderbolt display black screens
USB-C and Thunderbolt monitors have additional failure modes beyond traditional DisplayPort/HDMI. The Thunderbolt/USB4 protocol negotiates bandwidth for display alongside data transfer — when bandwidth negotiation fails or is disrupted by another device being connected to the Thunderbolt chain, the display briefly drops. This manifests as a brief black screen specifically when other USB-C or Thunderbolt devices are connected.
Ensure the Thunderbolt firmware is current (Windows Update sometimes delivers Thunderbolt firmware updates; also check the laptop manufacturer’s site). Connect the display directly to the Thunderbolt port nearest the CPU on the host device — some Thunderbolt implementations have port ordering where certain ports have priority access to the Thunderbolt controller.
Monitor-side failures
Some random black screens are the monitor itself — not the GPU, cable, or driver. Capacitors in older monitors fail and cause intermittent signal loss. The monitor-side panel backlight controller develops faults. Internal board connections loosen over years of thermal cycling.
The definitive test: use a completely different monitor (or TV via HDMI) connected to the same GPU port with the same cable. If black screens stop: the original monitor has a hardware issue. If they continue: the problem is the GPU, cable, or system rather than the specific monitor. This test takes 5 minutes to set up and definitively separates monitor-side from system-side causes.
Windows Fast Startup and display initialisation
Fast Startup’s partial kernel save occasionally includes stale display driver state. When Windows resumes from Fast Startup with a partially incorrect display driver state: the GPU initialises incorrectly and produces brief black screens during the first few minutes of use, particularly when switching between applications or changing display modes. Fully shutting down (disabling Fast Startup or using Shift+Shutdown) and rebooting — rather than a normal shutdown → power on — provides a clean display driver initialisation that resolves this specific pattern.
Frequency of black screen events and what they indicate
Brief black screens every few hours, stable during light use, more common during heavy GPU work → TDR or PSU issue. Brief black screens at consistent intervals regardless of GPU load → cable or adaptive sync issue. Single-monitor black screen that other monitors don’t share → that monitor or its cable. Full system brief black → GPU driver reset. Black screen that requires cable unplug/replug to resolve → signal negotiation failure at the connector level.
The pattern and context of each black screen event is the most informative data available. Noting the exact circumstances — what was running, which monitors affected, whether it self-recovered or required intervention — significantly narrows the fix without requiring extensive trial-and-error.
Refresh rate and resolution changes during game launch
Many games change the display resolution or refresh rate when launched. The brief black screen during this transition is normal — the monitor is reconfiguring for the new signal parameters. If the black screen lasts longer than 5 seconds or doesn’t recover: the game’s target resolution or refresh rate exceeds what the monitor supports, and Windows is waiting for the monitor to accept parameters it can’t handle. Setting the game’s resolution to match the monitor’s native resolution and a refresh rate the monitor is rated for eliminates this.
In NVIDIA or AMD control panel: set a custom maximum refresh rate that can be maintained, and ensure the game’s video settings match achievable parameters rather than trying to run at resolutions or refresh rates the hardware chain can’t sustain.
Increasing TDR delay as a temporary measure
On machines where the GPU driver recovers but takes long enough that Windows forces a TDR reset anyway, increasing the TDR delay threshold prevents the forced reset: HKEY_LOCAL_MACHINESystemCurrentControlSetControlGraphicsDrivers → add DWORD value “TdrDelay” = 8 (decimal, seconds). The default is 2 seconds. This gives the driver more time to recover before Windows forces the reset, which helps GPUs that occasionally take longer than 2 seconds to recover from a complex rendering operation.
This is a workaround, not a fix. The underlying cause (driver bug, hardware marginality, thermal issue) remains. Use it as temporary relief while finding the root cause rather than as a permanent solution — a GPU that consistently needs more than 2 seconds to recover is showing signs of instability that will worsen over time.
After addressing the most likely causes (cable, driver, adaptive sync, thermal), the black screen events should stop or dramatically reduce. For events that persist through all fixes: the correlation data from Event Viewer, the temperature logs from HWiNFO64, and the cross-device test (different monitor, different cable, different port) provide the specific information needed to either identify remaining hardware issues or escalate to manufacturer warranty support with evidence-backed reports.
A final note for laptop users: the “monitor going black” on a laptop is almost always the screen timeout or the Modern Standby screen-off feature rather than the GPU issues above. Laptop displays that go dark after inactivity are governed by Settings → System → Power → Screen and sleep → Turn off my screen after. This is the expected behaviour and isn’t a fault. The GPU TDR, cable, and adaptive sync issues above apply primarily to desktop setups with external monitors — though laptop external monitors connected via HDMI or USB-C can exhibit all the same problems as desktop monitors. Related: Windows 11 Second Monitor Not Detected.
