Pooled across ten randomized trials contributing safety data, oral NMN did not increase overall adverse events, serious adverse events, withdrawals for adverse events or system-specific adverse events against placebo, and did not raise ALT or AST [1]. For a supplement sold on this scale, that is an unusually orderly file.
It is worth knowing what produced that result before you rest on it — which harms the trials went looking for, how they looked, and who was never in the room. This page is about the harm endpoints. The doses themselves are laid out trial by trial in the NMN dosage page, and what NMN does or does not achieve is a separate question with a less flattering answer.
The panels the trials actually ran
The dedicated NMN safety study is the clearest example of a trial built to find something rather than to miss it. Thirty-one healthy adults aged 20 to 65 took 1,250 mg of β-NMN once daily for up to four weeks, and the investigators ran anthropometry, hematology, blood biochemistry, urinalysis and body composition analysis. Nothing changed beyond normal physiological variation, and there were no severe adverse events [2].
A twelve-week placebo-controlled trial in 30 healthy volunteers at 250 mg a day repeated the exercise with physiological and laboratory testing throughout, and reported no abnormalities and no obvious adverse effects while whole-blood NAD⁺ rose significantly [3]. The dose-ranging trial that randomized 80 healthy middle-aged adults to placebo, 300, 600 or 900 mg daily for 60 days states that no safety issues were found on adverse event monitoring, laboratory measures or clinical measures, and that all 80 completed without a protocol violation [4].
The pattern matters more than any single result. These are not studies that asked participants whether they felt fine; they drew blood, took urine and measured. When a category’s safety claims usually rest on customer reviews, a set of trials with named laboratory panels is a genuine point in NMN’s favor, and it is one of the few this site gets to make.
The only published percentage
Almost every NMN trial reports adverse events as an absence. One reports a rate, and it comes from an unexpected place: a single-arm, open-label phase 1/2 trial of low-dose oral NMN in adults with steroid-refractory or steroid-dependent immune thrombocytopenia. Twenty-five patients took 450 mg twice daily for two weeks. There were no dose-limiting toxicities and no treatment-related serious adverse events; mild treatment-related adverse events occurred in 12% of patients and non-severe grade 1 infections in 8%, with immunoglobulin levels remaining stable [5].
Hold two things about that at once. It is a small, unblinded study in people with an autoimmune blood disorder, so the population is not yours. And it is the only published figure putting a number on how often anything mild happens — which tells you how thin the reporting is elsewhere, not that NMN is unusually well characterized.
The liver question, specifically
“Is NMN hard on your liver” is the most common worry people arrive with, and it has a direct answer. The 2026 systematic review and meta-analysis searched six databases to May 2026, included fifteen randomized trials with ten contributing to the safety analyses, across doses of 250 to 2,000 mg a day and durations from 14 days to 24 weeks, and found no significant elevation of ALT or AST [1]. Individual trials that ran liver panels reported the same [2] [3].
What that establishes is narrow and real: in healthy and mostly metabolically ordinary adults, at supplement doses, for up to six months, liver enzymes did not move. What it does not establish is anything about someone whose liver is already compromised, because no published trial enrolled that person. The first study designed to produce longer safety data in a population with organ impairment — a phase 2a trial of pharmaceutical-grade microcrystalline β-NMN at 1,000 mg twice daily for 24 weeks in 156 adults with diabetic kidney disease — has published its protocol but not its results [6].
At the top of the studied range
The highest administration on record is 1,000 mg twice daily for 28 days in 30 overweight or obese adults aged 45 and over, randomized 2:1 against placebo. Adverse events were similar between groups. Circulating NAD and its metabolites rose substantially, and the changes that reached significance were cardiometabolic rather than adverse: body weight fell by a difference of −1.9 kg (95% CI −3.3 to −0.5, p = 0.008), diastolic blood pressure by −7.01 mmHg (95% CI −13.44 to −0.59, p = 0.034) and LDL cholesterol by −18.73 mg/dL (95% CI −31.85 to −5.60, p = 0.007). Insulin sensitivity, muscle strength, aerobic capacity and intra-abdominal fat did not change [7].
Two grams a day for a month, in the population most likely to be sold this, without a tolerability signal. That is the useful upper marker. It is also four weeks, in thirty people, which is the same sentence that ends every paragraph on this page.
The gaps that are not about the molecule
Duration. Twenty-four weeks is the longest exposure in the pooled evidence [1]. People buy NMN intending to take it for years. Nothing published describes year two.
Who was excluded. The trials enrolled healthy or overweight adults, generally from their twenties upward. Pregnancy, breastfeeding, childhood and established liver or kidney disease are not covered by any of them, so “well tolerated” carries no information about those situations.
Interactions. None of these studies set out to test NMN alongside prescription medication. The question is open rather than answered, which is why your medication list belongs in front of a clinician — the same reasoning that applies to the oral precursor family as a whole.
What is in the bottle. NMN’s status as a lawful dietary supplement ingredient in the United States is unresolved, which is a regulatory fact rather than a toxicological one. It does not make NMN less safe. It does mean the shelf around it is less settled than the one nicotinamide riboside sits on, which is part of why the choice between the two is not only a question about evidence.
The practical read
If you are a healthy adult taking a trial-range dose of a product whose label names NMN and a milligram figure, the published safety evidence is as reassuring as short trials can be, and it was collected properly. Take it with food if your stomach objects, tell a clinician the compound and the dose, and do not treat “no adverse events” from four weeks in thirty-one people as a statement about the next five years.
And keep the safety question downstream of the buying question. Most oral precursor products cost a small fraction of an injectable, and what NMN actually is is worth settling before you compare two bottles at all. Dose matters more than brand here, and dose is what the per-gram ranking makes visible.