The doses that appeared in the NAD+ precursor studies.

The short version

This page reports the NAD+ dosage figures that show up in published studies, and only as study context — it gives no instructions to take anything, and it is not medical advice. Almost every human number is a precursor dose, NMN or NR, because oral NAD+ itself is poorly absorbed (a precursor is a building block the body converts into NAD+). The pattern across trials is consistent: more precursor, more blood NAD+. Below are the doses, the routes researchers used, and how long NAD+ actually persists once it is infused versus supplemented — each tied to the study that measured it. Read every figure here as "this is what an experiment administered," never as "this is what you should take."

Doses used in NAD+ precursor studies

In human trials, NMN appeared at 250–900 mg/day orally, with 250 mg/day the most-replicated dose and up to 1200 mg/day studied; 600 mg/day was identified as optimal in the multicenter dose-ranging trial ^[3][1]. Nicotinamide riboside (NR) appeared commonly at 250–1000 mg/day orally, with up to 3000 mg/day tested for safety in a Parkinson's-disease study ^[4]. Nicotinamide (NAM) has been studied at 500 mg twice daily for skin-cancer chemoprevention. For IV NAD+, reported wellness/clinical infusion protocols run roughly 250–1000 mg per session over several hours, and a pharmacokinetic study used a continuous 3 µmol/min infusion over six hours.

One pattern is worth noting: the doses span a wide range — from 100 mg/day of NR at the low end to 3000 mg/day at the high end — and the trials that pushed highest were generally safety studies asking whether a dose was tolerable, not efficacy studies claiming it worked better. The most-replicated efficacy doses sit far lower (250 mg/day NMN; a few hundred mg/day NR), which is why a bigger number on a label is not a stronger claim from the literature.

These are observed study parameters. Reporting that a trial administered 250 mg/day of NMN is a description of that experiment — not a recommendation, and not a statement that any dose is safe or effective for any individual.

How long does infused or supplemented NAD+ persist?

The two routes behave very differently. Infused NAD+ is cleared from plasma fast: a pilot study found near-complete plasma removal within roughly the first two hours of infusion, consistent with the in-vitro finding that extracellular NAD is degraded by surface ecto-enzymes before cells can take it up intact ^[12]. Oral precursors are the opposite — they are absorbed and raise whole-blood NAD+ over days to weeks, and that elevation is sustained through chronic dosing in 8–12-week trials (NR's 22%/51%/142% rise was maintained across the full eight weeks; NMN's rise held at days 30 and 60) ^[4][3]. So "how long does NAD+ last" depends entirely on whether you mean a rapidly-cleared infusion or a precursor regimen that shifts the steady-state blood NAD+ level.

Why the numbers are precursor numbers

Read the dose figures above carefully and a pattern jumps out: nearly every human number is a precursor dose, not a dose of NAD+. That is not an accident of how trials were designed — it is forced by the biology. Intact NAD+ is a large, charged dinucleotide (MW 663.43 Da, C21H27N7O14P2) that cells do not freely import, and extracellular NAD is broken down by surface ecto-enzymes before any uptake ^[12]. So a study that wants to raise the NAD+ pool administers a building block — NMN at 250–900 mg/day ^[1][3] or NR at 100–3000 mg/day ^[4] — that the body converts into NAD+ through the salvage and related pathways. When you see "NAD+ dosage" discussed anywhere, the honest translation is usually "precursor dosage," and this digest keeps that translation explicit rather than letting an NMN figure masquerade as an NAD+ figure.

Routes studied, and how they differ

Oral capsules and powders of NMN, NR and nicotinamide carry the bulk of the controlled human evidence ^[1][3][4]. Intravenous NAD+ infusion is used in wellness settings but has limited controlled data — mostly pilot and retrospective — and is rapidly cleared from plasma ^[12]. Subcutaneous and intramuscular NAD+ injection is compounded with minimal peer-reviewed pharmacokinetic data. Sublingual, intranasal, topical and transdermal patch products are marketed with little controlled evidence; the human pharmacokinetic data cited here are for oral precursors, not patches.

The route choice changes the pharmacokinetics completely. An oral precursor is absorbed, metabolized into NAD+, and shifts the steady-state whole-blood NAD+ level over days and weeks ^[4][3]. An IV NAD+ infusion delivers the coenzyme directly into plasma but is then cleared within roughly two hours ^[12], so it produces a transient spike rather than a sustained shift — and because extracellular NAD is degraded at the cell surface, even that plasma NAD+ is not taken up intact ^[12]. These are mechanistically different interventions that happen to share the letters "NAD," which is exactly why the literature treats the oral-precursor evidence and the IV evidence as separate bodies of work.

Stability and the compounded-injectable risk

Stability is a practical footnote with a sharp edge. NAD+ and NMN are hygroscopic and degrade with heat and moisture, and reconstituted injectable NAD+ should be kept cold and protected from light. Compounded injectables carry contamination risk that oral capsules do not: an FDA Class I recall — the agency's most serious category — has been issued for a compounded NAD+ injection over elevated bacterial endotoxin. That recall is a documented quality event tied to the compounding of an injectable product, not an approved medicine with a manufacturing standard behind it. It is the single clearest reason the oral-versus-injectable distinction is not merely pharmacological but a matter of product quality, and why this page reports the IV route with that caveat attached every time.