NVIDIA Reflex: Should You Turn It On? (On vs On+Boost)
Published
On this page
Should you turn on NVIDIA Reflex? Yes — in supported games it is one of the few latency settings that genuinely, measurably reduces system latency, with essentially no downside. It removes the render queue so your input reaches the screen faster, and the gain is biggest exactly where it matters: when you are GPU-bound in a fast game. Turn it on.
Reflex strips the render queue and caps you below the ceiling — a real latency win, not a placebo.
What Reflex actually does
By default, the CPU can run ahead of the GPU, queuing several finished frames waiting to be rendered. That queue is latency: your input is captured into a frame that then sits in line. Reflex keeps the CPU and GPU tightly in step so frames aren’t stacked up waiting, which shortens the path from click to pixel.
It also dynamically caps your frame rate just below your refresh/G-Sync ceiling, keeping you inside the variable-refresh range — the same trick a manual FPS cap performs, done automatically. This is a real, understood mechanism, not a folklore tweak.
On vs On+Boost
The Reflex setting has two active modes. Here is what each adds:
| Mode | What it does | When it helps |
|---|---|---|
| Off | Normal render queue | Never the right choice in a supported game |
| On | Removes render queue + caps below ceiling | The main win — keeps it here for most play |
| On + Boost | Adds: raises GPU clocks sooner instead of idling | A small extra squeeze in CPU-bound moments; higher power draw |
Plain On delivers the large majority of the benefit. On + Boost keeps the GPU from down-clocking during CPU-bound frames, trimming a little more latency at the cost of higher power and heat. If you want every last millisecond and don’t mind the draw, use Boost; otherwise plain On is the clean default.
When the gain is biggest
Reflex helps most when you are GPU-bound — GPU usage pinned near 100% — because that is when the render queue would otherwise be deepest. In CPU-bound or heavily-capped scenarios the queue is already shallow, so there is less for Reflex to remove and the gain is smaller. The good news: it does no harm in those cases, so leaving it on is always safe.
How it interacts with your FPS cap
This trips people up, so be clear about it: in a Reflex game, Reflex is your FPS cap. It already limits your frame rate just below the G-Sync/refresh ceiling for latency reasons, so you do not need to stack a separate manual cap on top. If you want to understand the cap math for non-Reflex games, see Best FPS Cap for Low Latency.
Reflex also slots directly into the G-Sync + V-Sync + Reflex low-latency setup — it’s the piece that handles the cap and the render queue in that combo.
Reflex vs Low Latency Mode
If a game supports Reflex, use Reflex — it is implemented inside the game engine and is more effective than the driver-level NVIDIA Low Latency Mode (NULL). Low Latency Mode is the fallback for games without Reflex. For the full comparison, see NVIDIA Low Latency Mode: Ultra, On, or Off?.
Turning it on and verifying
Find the Reflex option in the game’s video/graphics settings and set it to On (or On + Boost). Then confirm the win rather than assuming it — measure your input lag with Reflex on and off in the same scene, or read NVIDIA’s PC Latency overlay directly. The millisecond difference is usually obvious, especially when you’re GPU-bound. AMD owners get the equivalent from AMD Anti-Lag 2.
Related guides
- NVIDIA Low Latency Mode: Ultra, On, or Off?
- G-Sync + V-Sync + Reflex: The Lowest-Latency Setup
- AMD Anti-Lag 2: Should You Use It?
- How to Minimize Input Delay for Competitive Gaming
- How to Measure Input Lag
Turn Reflex on in every game that supports it — plain On for the big, clean win, On + Boost if you want the last millisecond and accept higher power draw. Pair it with a fine Windows timer resolution via Tier1Timer and a correct G-Sync setup, and you have the genuine low-latency stack — no folklore required.
Frequently asked questions
Does NVIDIA Reflex actually reduce input lag?
Yes, genuinely — this is one of the few latency tweaks with a clear, measurable mechanism. Reflex removes the render queue so the CPU stops preparing frames far ahead of the GPU, which cuts the time between your input and the frame that shows it. The benefit is largest when you are GPU-bound (GPU at near 100% usage), where the queue would otherwise be deepest.
What is the difference between Reflex On and On+Boost?
Reflex On removes the render queue and dynamically caps your frame rate just below your refresh/G-Sync ceiling. On+Boost adds a second behavior: it raises GPU clocks sooner instead of letting them idle, shaving a little extra latency in CPU-bound moments at the cost of higher power draw. For most players, plain On is the bigger, cleaner win; Boost is a small extra squeeze.
Should I turn Reflex on if I'm not GPU-bound?
There is little downside to leaving it on. When you are CPU-bound or already capped well below your GPU's ceiling, the render queue is shallow so Reflex has less to remove — the gain is small. But it does no harm in those scenarios, and the moment you become GPU-bound it starts working, so 'on' is a safe default in any supported game.
Do I still need an FPS cap if Reflex is on?
In a Reflex game, no — Reflex automatically caps your frame rate just below your G-Sync/refresh ceiling, which is exactly what a manual cap does for latency. You can let Reflex handle it. If a game lacks Reflex, set a manual cap a few frames below your refresh rate yourself.
Which games support NVIDIA Reflex?
Most major competitive titles support it — Valorant, CS2, Apex Legends, Fortnite, Overwatch 2, Rainbow Six Siege, and many more — plus a growing list of single-player games. It requires a GeForce GTX 900-series or newer GPU. If you see a Reflex option in the game's settings, turn it on; if there's no option, the game doesn't support it and you fall back to a manual FPS cap.