If you have spent any time on the wellness side of TikTok in the past two years, you have seen the video: a small jar of black, sticky resin, a spoon pulling a long viscous strand out of it, and a voiceover about ancient Himalayan vitality. The substance is shilajit, and it has become one of the breakout supplement categories of the decade. Market analysts put the global category at roughly $207 million in 2025 and project it past $435 million by 2034.[8]
Those projections come from commercial market-research firms and should be read as estimates rather than audited figures. But the direction is not in doubt: shilajit went from an obscure Ayurvedic preparation to a mainstream men's-health product in about thirty-six months, largely on the strength of one claim — that it raises testosterone.
So we did the boring thing. We searched PubMed for every randomised controlled trial ever published on shilajit in humans. The search returns seven papers.[9] Here is what all seven of them found.
What shilajit actually is
Shilajit is not a herb, a root or an extract of anything living. It is an exudate: a blackish-brown organic-mineral material that seeps from rock crevices in high mountain ranges, formed over very long periods from compressed plant and microbial remains. The same substance appears under different names across a wide geography — mumijo, mumie, moomiyo, mineral pitch.
Chemically it is a mixture, not a compound. The two constituent classes that get named on labels are fulvic acids and dibenzo-α-pyrones. Fulvic acid is worth pausing on: it has no single molecular formula, because it is not a single molecule. It is an operationally defined family of humic substances of varying size and structure, which is precisely why standardising a shilajit product is difficult. The dibenzo-α-pyrone core (C13H8O2) is a better-defined marker, and readers who have followed our coverage of urolithin A may recognise the scaffold — urolithins share the same dibenzopyranone skeleton.
The trials below almost all used purified, standardised extracts with declared fulvic acid content. That is not the same product as raw resin sold by weight, and the distinction matters more here than for almost any other supplement.
All seven trials, in one table
| Year / journal | Who | Dose & duration | Main finding |
|---|---|---|---|
| 2016 Andrologia | Healthy men, 45–55 | 250 mg twice daily, 90 days | Total testosterone, free testosterone and DHEAS up vs placebo (p < 0.05); LH and FSH unchanged[1] |
| 2019 J Int Soc Sports Nutr | 63 active men, ~21 yrs | 250 or 500 mg/day, 8 weeks | Less strength decline after fatigue at 500 mg — but only in a pre-specified upper-half subgroup[2] |
| 2019 J Am Coll Nutr | Healthy adult women | 125 or 250 mg twice daily, 14 weeks | Improved skin microperfusion at the higher dose; skin genes for blood-vessel growth and extracellular matrix upregulated[3] |
| 2020 J Altern Complement Med | 160 tibia fracture patients | 1000 mg/day, 28 days | Mean time to bone union 129 days vs 153 on placebo (p < 0.049); no difference in adverse effects[4] |
| 2022 Phytomedicine | 60 postmenopausal women with osteopenia | 250 or 500 mg/day, 48 weeks | Dose-dependent preservation of spine and femoral neck bone density; bone-turnover, oxidative-stress and hsCRP markers improved (p < 0.001)[5] |
| 2022 J Diet Suppl | 35 trained men, ~21 yrs | 500 or 1000 mg/day, 8 weeks | Serum pro-C1α1, a type 1 collagen synthesis marker, roughly doubled; placebo unchanged[6] |
| 2025 Nutrients | 109 completers with metabolic syndrome risk | 6–12 mg inside a combination product, 12 weeks | Some cardiometabolic improvements — but shilajit was a minor component alongside chromium and fruit extract[7] |
The testosterone trial everyone cites
Reference 1 is the paper underneath the entire marketing category. It is a genuine randomised, double-blind, placebo-controlled study, published in a peer-reviewed andrology journal, and it found what it says it found: 90 days of purified shilajit at 250 mg twice daily raised total testosterone, free testosterone and DHEAS relative to placebo in healthy men aged 45 to 55, while the gonadotropins LH and FSH stayed level.[1]
Three things are worth saying about it plainly.
First, it is one trial. In a field where replication is the whole game, an unreplicated result from 2016 that has attracted a nine-figure market and no confirmatory study in a decade is a conspicuous gap, not a foundation.
Second, the "20 percent testosterone increase" that circulates in product copy and video captions is not a figure the published abstract states. It comes from secondary write-ups. When a number has travelled that far from its source, treat it as marketing until you can see it in the paper.
Third, the participants were men aged 45 to 55 — an age band where testosterone is typically drifting downward. Whether the same effect appears in the 22-year-old buying the resin off a TikTok Shop link is untested, and there is a reasonable physiological argument that it would not.
The best trial is the one nobody talks about
Judged purely on design, the strongest study in the set is reference 5: 60 postmenopausal women with osteopenia, randomised across three arms, followed for 48 weeks with bone density measured by scan at weeks 0, 24 and 48 and a full panel of turnover and inflammation markers along the way.[5]
Bone density fell in the placebo group, as expected, and that decline was attenuated dose-dependently in both supplemented groups. Markers moved coherently in the same direction — bone resorption markers down, osteoprotegerin up, malondialdehyde down, glutathione up, hsCRP down. A dose-response relationship plus a mechanistically consistent biomarker pattern over nearly a year is a much better piece of evidence than a single 90-day hormone readout.
It is also, tellingly, not what shilajit is being sold for. The bone and connective-tissue signal is the most repeated finding across this literature — it shows up again in the fracture-healing trial[4] and in the collagen-synthesis marker study[6], which is a more interesting thread than the testosterone story and connects to what we have written about collagen supplements and skin.
Where the evidence gets thin
Now the caveats, which are substantial.
The sample sizes are small and the participants are young men. Two of the seven trials recruited a combined 98 recreationally active men in their early twenties and ran for eight weeks. The strength result[2] was significant only within an upper-half subgroup split after the fact by baseline values, not in the group as a whole — a pattern that frequently fails to survive replication. The collagen-marker result[6] looks dramatic until you read the variability: the high-dose group averaged 113.1 ng/mL after supplementation with a standard deviation of ±78.7, meaning individual responses were scattered enormously.
One trial can't tell you anything about shilajit at all. The 2025 Nutrients study[7] is the largest and best-resourced on the list, but shilajit appeared only as a 6–12 mg component inside a combination product built around chromium and Phyllanthus emblica extract, given alongside a 12-week supervised exercise and diet programme. Whatever improved, there is no way to attribute it to the shilajit.
Much of the evidence has one commercial parent. Four of the seven trials used PrimaVie®, a single branded standardised extract. Industry-funded research is normal and not automatically suspect — someone has to pay for trials — but a literature where most positive findings trace to one supplier's ingredient is a literature that has not been independently stress-tested.
Nobody has studied shilajit for aging. This is the one that matters most for our readers. There is no human trial of shilajit against any aging endpoint — not mortality, not epigenetic age, not incident disease. The antioxidant and anti-inflammatory markers in the osteopenia trial are the closest thing, and biomarkers are not outcomes. Anyone selling shilajit as a longevity compound is extrapolating well past the data, which is the recurring problem across most of the supplements marketed for longevity.
The heavy metals question, answered properly
The standard criticism of shilajit is that a substance scraped off a mountain is likely to be loaded with lead, arsenic and mercury. It is a fair concern, and until recently the published data on it was poor.
In June 2026, researchers in Poland, France and Slovakia published the most thorough elemental characterisation to date: twelve shilajit samples from diverse geographical origins, analysed by inductively coupled plasma mass spectrometry alongside field-flow fractionation and size-exclusion chromatography.[10]
The result is more interesting than either camp wants. Toxic elements — cobalt, nickel, lead, cadmium, mercury — were detectable but below established toxicity thresholds. Rare earth elements were present in distributions following natural geochemical patterns, which the authors read as environmental uptake during formation rather than contamination or adulteration. Calcium was the most abundant element, and most metals were bound in low-molecular-weight complexes under 10 kDa.
So the panic version is overstated. But note what that study does and does not license you to conclude. Twelve samples, characterised in a laboratory, tell you about the geology of shilajit as a class. They tell you nothing about the specific unbranded jar shipped to you by a seller with no certificate of analysis. Composition varies by source rock, and the supplement market's quality-control problem is a supply-chain problem, not a geochemistry one. The practical rule is unchanged: buy only products with batch-specific third-party testing you can actually read.
The honest summary
Shilajit is not snake oil. Seven randomised trials is more human evidence than a great many supplements on the shelf can claim, the bone and connective-tissue findings are reasonably consistent, the 48-week osteopenia trial is a real piece of work, and the safety signal across the published studies is unremarkable — the fracture trial specifically found no difference in adverse effects versus placebo.[4]
But the gap between what has been shown and what is being sold is wide. The testosterone claim carrying the market rests on a single unreplicated study in middle-aged men, quoted with a number that isn't in the paper. The aging claims rest on nothing. And the most defensible findings are about bone density in postmenopausal women, which is not who the advertising is aimed at.
If you want the version of this decision that applies to every supplement: the evidence bar for a compound worth taking is replication by independent groups against outcomes that matter. Shilajit clears the first half of that bar in one narrow domain and none of it in the domain it is famous for. That is a reasonable thing to keep an eye on, and a poor thing to build a routine around — particularly compared with the interventions in our list of drugs and compounds that may slow aging, where the human data actually exists.
Common questions
Does shilajit increase testosterone?
One randomised, double-blind, placebo-controlled trial has tested this — 250 mg twice daily for 90 days in healthy men aged 45 to 55, reporting significant rises in total testosterone, free testosterone and DHEAS with LH and FSH unchanged.[1] It has never been independently replicated, and the "20 percent" figure in product copy is not stated in the published abstract. One trial is a reason to watch, not a settled fact.
Is shilajit full of heavy metals?
The best available analysis, published June 2026, ran twelve samples from different origins through ICP-MS and found toxic metals present but below established toxicity thresholds, with rare earth patterns consistent with natural formation rather than adulteration.[10] That is reassuring about shilajit as a material. It is not a guarantee about an untested jar from an unknown seller — insist on a batch-specific certificate of analysis.
What dose did the trials use?
There is no established dose, because no dose-finding study exists. Published trials cluster between 250 mg and 1000 mg daily of a purified, standardised extract, most commonly 500 mg split into two doses, over 8 to 48 weeks. Raw resin is not equivalent to a standardised extract. Anyone on prescription medication, pregnant or breastfeeding, or with kidney disease, gout or an iron-overload condition should talk to a doctor before starting.