Spirulina is a cyanobacterium — a photosynthetic bacterium — grown in shallow ponds and dried into a deep blue-green powder. It runs 60 to 70 percent protein by dry weight, and the human evidence behind it is real but thinner than the marketing implies. This article covers what is actually in it, what the studies found and how good those studies are, why the common claim that spirulina is a vegan source of vitamin B12 is false, and why cultivation and batch testing matter more than a romantic origin story. Some of this is less flattering than the label copy.
It is a bacterium, not an alga
Commercial spirulina is Arthrospira platensis or Arthrospira maxima, a photosynthetic prokaryote — closer to a bacterium than to seaweed, despite sitting in the same aisle. The NIH LiverTox monograph classifies it that way, and the distinction is not pedantry.
The looser term “blue-green algae” covers a broader regulatory category, including wild-harvested Aphanizomenon flos-aquae, which has a different contamination profile. A label that will not name the species will not tell you which one you are buying.
What is actually in 10 grams
USDA composition data for dried spirulina, scaled to a 10 g serving, gives about 5.7 to 6.1 g of protein, about 2.9 mg of iron, 0.24 mg of thiamin, 0.37 mg of riboflavin, 1.3 mg of niacin, 0.6 mg of copper and about 105 mg of sodium (full USDA entry). The protein is complete, with all nine essential amino acids.
The other component worth naming is C-phycocyanin, the pigment that makes the powder blue rather than green. A 2026 review in Frontiers in Nutrition puts it at around 47 percent of total protein, roughly 2 to 3 g per 10 g serving. It is the compound most of the antioxidant research is built around.
One caveat on all of these numbers. Published carotenoid figures differ between USDA and review compilations by orders of magnitude, and iron values in that same review span more than twofold, because strain, light and cultivation conditions genuinely move them. Any per-serving figure should come from the certificate of analysis for the actual batch, not a database average.
The B12 problem, stated plainly
Spirulina is not a source of vitamin B12, and we are not going to print that it is. The B12 that shows up on an assay is overwhelmingly pseudovitamin B12, or pseudocobalamin — a structural analogue that human B12-dependent enzymes cannot use.
Watanabe and colleagues analyzed spirulina tablets and found pseudocobalamin was the predominant cobamide present, concluding the product “is not suitable for use as a B12 source, especially in vegetarians.” The mechanism is straightforward: cyanobacteria run on a different cobamide variant than human metabolism requires. USDA lists B12 in spirulina as not detected.
This matters most for exactly the people the claim targets. Anyone moving to a plant-based diet needs a genuine B12 source — a supplement or fortified food supplying cyanocobalamin or methylcobalamin. Relying on spirulina for it produces a shortfall slowly and quietly.
What the human trials found
Blood lipids
This is the strongest signal in the literature. Rahnama and colleagues published a GRADE-assessed dose-response meta-analysis in Pharmacological Research in 2023, pooling 20 randomized controlled trials and 1,076 participants. LDL cholesterol, total cholesterol and triglycerides each fell relative to control, with a standardized mean difference of about −0.6 (95% CI −0.9 to −0.2), and HDL cholesterol rose slightly (SMD +0.3, CI 0.0 to 0.6).
Blood pressure
Machowiec and colleagues, writing in Nutrients in 2021, pooled five randomized trials (230 subjects) for systolic and four (214 subjects) for diastolic readings, at 1 to 8 g per day over 2 to 12 weeks. Systolic readings were 4.59 mmHg lower than control (CI −8.20 to −0.99), diastolic 7.02 mmHg lower (CI −8.86 to −5.18) once one high-heterogeneity trial was excluded. The effect was concentrated in participants who already had hypertension (−9.18 mmHg) rather than those with normal readings (−2.26 mmHg). The authors flagged the small number of trials and the absence of any publication-bias assessment.
Exercise and recovery
Liu and colleagues pooled 22 randomized trials and 822 participants across seaweed and microalgae in Nutrients in 2026. Across all species the numbers looked good — VO₂max SMD 0.88, time to exhaustion 1.06, creatine kinase −0.78. Spirulina’s own subgroup did not carry that. Its aerobic effect was weak (SMD 0.23), its recovery effect modest (−0.39), and heterogeneity was high throughout.
Where the evidence is weak
NIH’s assessment is that the evidence is insufficient to recommend spirulina for any human condition. That is the honest headline, and it sits above every finding in the section before this one. The nulls belong in the record too: in type 2 diabetes, pooled trials found fasting glucose fell while HbA1c did not move — and HbA1c is the more meaningful of the two.
The structural problems repeat. Trials are small, often under 100 participants, and short — mostly 4 to 12 weeks, which tells you nothing about a year of use. They cluster in a handful of research groups and countries, are frequently open-label rather than blinded, and industry funding is common; an umbrella review of the field documents the pattern. Some popular claims rest on studies with no control group at all.
None of that makes the lipid findings fake. It makes them provisional. A meta-analysis of 20 small short trials is not the same class of evidence as one large long one.
Trial doses versus what a scoop delivers
Trial doses cluster between 1 and 8 g per day, most commonly 2 to 4.5 g, run over 4 to 12 weeks. LiverTox cites 4.5 to 6 g per day as typical supplement use, and standalone spirulina products are usually sold at around 3 g per serving.
| Source | Daily spirulina |
|---|---|
| Range across published trials | 1–8 g |
| Most common trial dose | 2–4.5 g |
| LiverTox, typical supplement use | 4.5–6 g |
| Standalone spirulina, usual serving | about 3 g |
| Spirulina share of a blended greens scoop | often under 2 g |
Here is the part the category tends not to say out loud. A blended greens scoop is usually 8 to 12 g in total, divided across dozens of ingredients, so the spirulina share commonly falls between a fraction of a gram and about 2 g — frequently below what the trials used. A blend listing 30 ingredients and one total scoop weight tells you almost nothing about the dose of any single one.
So the practical position: a greens powder is a broad nutritional base, not a delivery vehicle for trial-level spirulina. Anyone who wants the intake used in the research should look at standalone spirulina alongside a blend, and should expect per-ingredient weights on the label rather than a proprietary blend. Our guide to reading a greens powder label covers which terms are useful and which are evasions.
Cultivation and testing beat the origin story
Spirulina does not produce microcystins. The contamination risk comes from co-culture — toxin-producing cyanobacteria such as Microcystis aeruginosa establishing themselves in an open or poorly controlled pond. That risk is not theoretical. A 2020 survey in Toxins tested 10 commercial microalgae supplements and found three of ten positive for microcystin-LR or anatoxin-a, with lead detectable in six of ten, though within limits. WHO’s tolerable microcystin intake is around 0.24 µg per day for a 60 kg adult, and Health Canada caps microcystin-LR in raw material at 1 ppm.
On the toxin that draws the most alarmed coverage: France’s food safety agency ANSES reports that BMAA has not been detected in marketed spirulina batches. It is an open question for the wider algae category, not an established spirulina hazard. ANSES did log 49 adverse-event reports between 2009 and 2017 — angioedema, anaphylaxis, digestive complaints, myositis, and kidney and liver effects among them — and it recommends favoring “supply channels with the best oversight by public authorities.” LiverTox rates spirulina category D: a possible but very rare cause of clinically apparent liver injury, with none observed in controlled trials.
Kona on Hawaii’s Big Island is a genuinely good place to grow it. Producers there describe it as the sunniest coastal location in the US with a 12-month season, and describe shallow paddle-wheel-stirred ponds fed by rainforest-aquifer well water and deep ocean water pumped from 2,000 feet, biomass screened and triple-rinsed, then dried under low oxygen to limit oxidation. They also state a certified biosecure zone, 15-plus test categories, and microcystin testing that has never returned a positive.
Those are supplier statements, and worth reporting as such rather than as verified fact. The defensible claim, and the one ANSES and the Toxins survey both support, is narrower and more useful: controlled cultivation plus batch third-party testing separates safe product from risky product. Latitude is a nice detail; a current certificate of analysis is the thing that tells you something — the same standard that applies to chlorella and how it is processed.
Who should not take it
- Phenylketonuria — avoid entirely. Spirulina is high in phenylalanine, with USDA listing 2,771 mg per 100 g, roughly 277 mg per 10 g serving. ANSES names PKU explicitly.
- Autoimmune conditions or immunosuppressant therapy — caution. Spirulina has documented immunostimulatory activity. Get clinician sign-off first.
- Pregnancy and breastfeeding — caution. LactMed reports no data on transfer into breastmilk or on infant safety, and flags uncontrolled production as the main concern.
- Anticoagulants — caution. Check the interaction with a prescriber before starting.
- Allergy-prone, or a history of muscle or liver problems. ANSES advises against use, citing reports of angioedema, anaphylaxis and myositis.
- Minor effects reported by others include diarrhea, bloating, headache and flushing, usually early on.
Common questions
Is spirulina a good source of B12 for vegans?
No. Its B12 is overwhelmingly pseudocobalamin, an analogue human enzymes cannot use, and USDA lists active B12 as not detected. Anyone eating plant-based needs a supplement or fortified food supplying cyanocobalamin or methylcobalamin. A label calling spirulina a vegan B12 source is wrong.
How much spirulina do the studies use?
Most trials used 2 to 4.5 g per day, within an overall range of 1 to 8 g, over 4 to 12 weeks. A typical blended greens scoop contains considerably less spirulina than that. If matching the research dose matters to you, a standalone spirulina product is the honest way to do it.
Does spirulina detox heavy metals?
No. There is no human evidence that a spirulina serving removes heavy metals from the body, and the phrase describes nothing measurable. The heavy-metal question that matters runs the other way: whether the powder itself was tested for lead and cadmium before it was sold.
Is Hawaiian spirulina better?
Kona has real advantages as a growing site, but those are production conditions, not a guarantee about a specific bag. What separates a good batch from a bad one is a closed, well-managed cultivation system and third-party testing of that batch for microcystins and heavy metals. Ask for the certificate rather than the postcard.
Spirulina is a dense complete protein with an interesting pigment and a provisional evidence base built on small short trials. It is not a B12 source, it is not a detox agent, and a scoop of any blend rarely delivers what the studies used. What it is, grown and tested properly, is a well-characterized ingredient that earns its place in a broader mix.
That is the standard we hold our own ingredients to, and you can see the full formulation and testing approach at Greens Plus.
About the Author
