Rhubarb leaf blades are poisonous and must never be eaten - only the leafstalks (the petioles) are edible. Harvest the stalks and cut the leaves off.
Well established · suboptimal if ignored → advise
How it works
The leaf blades hold a toxic concentration of oxalic acid - and, per NDSU, anthraquinone glycosides - which the edible petioles carry only in trace amounts. Oxalic acid binds calcium and can damage kidney function; eating the blades causes cramping, nausea and, in quantity, can be fatal. The stalks are safe because the toxins concentrate in the leaf, not the petiole.
What to do about it
Cut the leaf blades off and discard them; eat only the stalks. Never cook or eat the leaves.
How much it matters: Not quantified as a threshold dose. The sources give the hazard and the compound, and NDSU's plant-toolbox entry rates the severity "Medium" and notes it is "TOXIC ONLY IF LARGE QUANTITIES EATEN" - a qualitative caution, not a stated gram figure, so none is invented here.
The evidence
Evidence: confirmed against 5 sources in the research-based literature.
the 5 sources
UW-Madison / Wisconsin Master Gardener (Susan Mahr), 'Rhubarb, Rheum rhabarbarum', mastergardener.extension.wisc.edu/files/2018/04/rhubarb.pdf (read 2026-08-05): 'The leaves, however, contain a toxic amount of oxalic acid and are poisonous. The ensuing cramps, nausea and sometimes death from ingestion suppressed interest in the plant for about two hundred years.'
NDSU Extension, 'Dakota Gardener: Is Frost-Damaged Rhubarb Toxic?', ag.ndsu.edu/news/columns/dakota-gardener/dakota-gardener-is-frost-damaged-rhubarb-toxic (read 2026-08-05), names BOTH compounds and the edible/toxic split outright: 'The leaves contain oxalic acid which can damage kidney function'; 'recent scientific studies show the presence of other compounds such as anthraquinone glycosides that may contribute to the toxicity of the leaves'; 'Rhubarb stems (petioles) which resemble celery stalks are the edible portion of the plant because they contain only trace amounts of toxic compounds.' The SAME page is why the frost->toxin belief is excluded: 'A hard freeze is widely believed to cause the movement of toxic compounds from the leaves into the stems' ... 'I could not find any scientific studies to confirm or disprove this belief.'
U. of Illinois Extension (Good Growing), 'To rhubarb or not to rhubarb?', extension.illinois.edu/blogs/good-growing/2023-06-08-rhubarb-or-not-rhubarb (read 2026-08-05): 'Do not consume leaf blades, they contain large amounts of oxalic acid which can limit the body's ability to take up calcium.'
UMN Extension, 'Growing rhubarb in home gardens', extension.umn.edu/vegetables/growing-rhubarb (read 2026-08-05): 'Rhubarb leaves are toxic. Do not eat them.'
NC State Extension Gardener Plant Toolbox, 'Rheum rhabarbarum', plants.ces.ncsu.edu/plants/rheum-rhabarbarum/ (read 2026-08-05): Poison Severity 'Medium'; Poison Parts 'Leaves'; Toxic Principle 'Anthraquinone glycosides and soluble oxalates'; Poison Symptoms 'TOXIC ONLY IF LARGE QUANTITIES EATEN. Abdominal cramps, burning of mouth and throat, headache, weakness, nausea and vomiting, coma; may be fatal'.
Where this rule stops
RHUBARB ONLY, scoped by applies_to.species like another rule here - it reads no species field, because "the leaves are toxic" is a fact about a named crop, not a property the corpus measures. It is the first toxicity rule in the corpus, and it exists because the alternative did not fit: the hazard entity (juglans_nigra) models SPATIAL allelopathy - an effect radius and a list of sensitive neighbours - while this is INGESTION harm to the gardener with no radius and no neighbour, so a rule with a mechanism is the honest home for it (maintainer ruling, 2026-08-05). THE SEVERITY AXIS MEASURES GARDEN OUTCOME (how bad for the planting if ignored); this is a USE-SAFETY advisory, so it sits at `advise` and carries its seriousness in the claim wording, never in a severity the app does not show (D-105). Two claims are DELIBERATELY NOT made: LIVESTOCK toxicity, asserted only by non-extension secondary sources, so it waits for an extension-grade source that names it; and the FROST->TOXIN belief that a hard freeze moves oxalic acid into the stalks - Illinois states it, NDSU says it is scientifically unconfirmed, so the corpus does not carry a contested claim as fact.