NAD+ Dosing for Cognitive Benefits: What the Evidence Actually Supports

Search for an NAD+ dose and you'll get confident numbers: 300 mg of NMN, 500 mg of nicotinamide riboside, "1 gram twice daily" for serious results. The figures look precise, which makes them feel evidence-based. But there's an uncomfortable fact underneath the precision: no dose of any NAD+ precursor has been shown, in a controlled trial, to improve cognition, focus, or mental clarity in people. The doses that exist were chosen to move a blood marker or to test safety — not because a study found they made anyone think more clearly. This is a guide to what the dosing evidence really says, and to the gap between the numbers on a supplement label and the outcome you actually care about.

The bottom line up front

There is no "cognitive dose" of NAD+. The human trials that picked specific doses did so to study pharmacokinetics (how much NAD+ ends up in your blood) or safety — not cognition. When those same precursors were tested against thinking, the results were flat: NAD+ went up, cognition didn't budge⁵⁶. So any "take X mg for sharper focus" recommendation is extrapolating from a biomarker, not reporting a cognitive result.

Two implications follow, and they matter before you spend a cent. First, brain fog is a symptom with real, often-treatable causes — sleep debt, thyroid disease, B12 deficiency, medications, post-viral illness — and chasing a supplement dose skips the step that actually fixes most cases. We lay out that order of operations in What Actually Causes Brain Fog?. Second, NAD+ precursors are supplements, not drugs: they aren't FDA-approved to treat any cognitive condition, and the "right dose" is genuinely unsettled.

Why a dose can raise NAD+ without doing anything for your mind

It helps to separate two questions that marketing deliberately blurs:

1. 1.Does this dose raise NAD+? Often yes — and that's measurable.
2. 2.Does raising NAD+ at that dose improve cognition? That's the question the trials keep failing to answer in the affirmative.

On the first question, the dose-response data are real. An open-label pharmacokinetic study of nicotinamide riboside showed it raised the blood NAD+ metabolome in healthy adults in a dose-dependent way⁴. For NMN, a single-dose study in healthy men found oral doses of 100, 250, and 500 mg were safely metabolized with no serious problems³, and a 250 mg/day regimen in prediabetic women reliably affected downstream biology². So precursors do move the marker, and bigger doses generally move it more. That's the half of the story the labels tell.

The second question is where it falls apart. In a randomized, placebo-controlled trial of nicotinamide riboside in older adults with mild cognitive impairment, NAD+ rose — and cognition did not improve versus placebo⁵. A 2025 randomized trial in long-COVID, a population defined partly by brain fog, found NR raised NAD+ roughly 3-fold yet produced no significant benefit for cognition, fatigue, sleep, or mood⁶. A review focused specifically on NR for cognitive impairment lands on the same cautious verdict: biologically interesting, not yet shown to protect thinking in people⁹. The marker moved at these doses; the minds didn't. That's the entire reason "dose for cognition" is a category error right now — you can't titrate to an outcome no dose has been shown to produce.

What the dose-ranging trials actually tested

The most useful dosing evidence comes from trials that deliberately compared different doses. They're worth reading honestly, because they're the closest thing to a real answer — and they still don't get you to cognition.

The strongest example is a randomized, multicenter, double-blind, placebo-controlled, dose-dependent trial of NMN in healthy middle-aged adults that compared 300, 600, and 900 mg/day¹. It found higher blood NAD+ with higher doses and reported the regimen was well tolerated. Crucially, its primary readouts were a walking/endurance measure and blood NAD+ — not a validated cognitive battery. So even the best dose-ranging study in the field tells you which dose raises NAD+ more, not which dose helps you think.

For nicotinamide riboside, the dose-response PK study covered the same ground on the absorption side — more NR, more NAD+ in the blood⁴ — without claiming a cognitive endpoint. And on the high end, the NR-SAFE trial pushed to 3000 mg/day (a deliberately high dose) primarily to test safety, reporting it was generally well tolerated⁷. That's a tolerability ceiling, not a cognitive recommendation. Notice the pattern: every dose number in this literature is anchored to a biomarker or a safety endpoint. None is anchored to a cognitive result.

For the deeper dive on precursor-specific brain data, see NMN & NR for Brain Health: What the Studies Show.

"Higher dose, sharper mind" is the wrong mental model

It's tempting to assume that if a low dose nudges NAD+, a high dose will finally unlock the cognitive benefit. The trials argue against that:

• •The cognition trials already used substantial doses and still came up empty. The MCI and long-COVID studies weren't underdosed — they raised NAD+ meaningfully and saw no cognitive signal⁵⁶. If a real effect were hiding behind dose, you'd expect at least a hint at those exposures.
• •More NAD+ in blood is not more NAD+ where it might matter. Raising the circulating marker doesn't establish that brain tissue exposure changes in a way that improves function — and no human cognitive endpoint has been tied to any dose.
• •Safety is reasonably reassuring, but that's a separate axis. A systematic review of NAD-based interventions across conditions found them generally well tolerated while stressing that evidence for benefit is inconsistent and the strongest claims outrun the data⁸; high-dose NR was likewise generally safe in a randomized trial⁷. "You can probably take more safely" is not "more will help you think." Chasing dose for a benefit no dose has produced is cost and risk without a payoff to point to.

What about IV and other routes that promise to make dose moot? The only human IV-NAD+ study measured plasma chemistry during the infusion, not cognition, and didn't establish a cognitive dose either¹⁰. We unpack that route in NAD+ IV for Mental Clarity: Is It Worth It?.

If you're set on trying a precursor anyway

This isn't a "never" — NAD+ precursors are generally well tolerated and the choice is yours. It's a "go in with clear eyes." A few honest guardrails:

• •Anchor to the trials, not the marketing. The doses with the most human data are roughly 250–900 mg/day for NMN¹²³ and up to about 1000 mg/day (and higher in safety studies) for NR⁴⁷. These are studied doses, not proven-for-cognition doses — a distinction the label will never make for you.
• •More isn't a strategy. Since no dose has been shown to help cognition, escalating to chase an effect just raises cost and exposure without an evidence basis.
• •Set a real expectation and a stop date. Any felt "clarity" in the first weeks is just as likely placebo, better sleep, or hydration. Decide in advance what would count as a benefit and when you'll quit if it doesn't appear.
• •Talk to a clinician if you take medications or have a health condition. Supplement-drug considerations and underlying conditions don't show up on a label.

What's actually worth "dosing" for a clearer head

If the goal is real cognitive improvement, the evidence points to levers that aren't NAD+ at all. Sleep is the single best-supported driver of attention and processing speed; reviews document dose-dependent cognitive decline with sleep loss¹¹. Common reversible medical causes are routinely missed: thyroid dysfunction independently affects mood and cognition¹², and vitamin B12 deficiency is a classic, correctable cause of fatigue and cognitive slowing¹³. Among consumables people actually take, omega-3 has the most credible — still modest — general-cognition evidence in adults without dementia, and notably it does show a dose-response¹⁴. None is a miracle. But each has more behind its "dose" than any NAD+ precursor does for thinking.

The honest verdict

There is no NAD+ dose proven to improve cognition, because no precursor at any dose has cleared that bar in a controlled trial. The dose numbers you see — 300/600/900 mg of NMN, up to a gram or more of NR — come from studies that measured blood NAD+ or safety, then saw cognition stay flat when anyone bothered to test it. Treat precursors as supplements, not cognitive drugs; if you try one, anchor to studied doses, don't escalate to chase a benefit that hasn't been demonstrated, and put your real effort into sleep and a basic medical workup. For the full picture of what NAD+ does and doesn't do for foggy thinking, see our pillar review, NAD+, Brain Fog & Focus: What the Evidence Shows — and if you still want to weigh cognitive-energy products against that evidence bar, see our best cognitive-energy picks.