Every LabelCheck every label, every time.

← Cross-reactivity guides

Legume cross-reactivity

By Symon, Every Label ·

If you or your child is allergic to a legume, almost always peanut, the fear is that the whole legume aisle is now off limits: soy, peas, lentils, chickpeas, beans, all of it. The honest answer is reassuring. A legume allergy does not mean you must avoid all legumes. Most people with a peanut allergy tolerate the other common legumes, because cross-reactivity between peanut and soy, peas, lentils, and chickpeas is low, even though positive tests across legumes are common. A positive test is co-sensitization, not the same as a reaction. Confirm any specific food with your allergist before introducing it.

There are real exceptions, and the page is honest about them: lupin is the legume that genuinely cross-reacts with peanut, fenugreek is the second, and within the bean-and-pea group lentil, chickpea, and pea genuinely travel together. This page is the category map. It is the deep version that each legume profile’s cross-reactivity summary links out to: which legumes cross-react, which do not, and the protein-level reason for the difference. Where a claim is a verified cross-reactivity fact, it is drawn from the project’s cross-reactivity floor. Where it is allergen-specific clinical detail, it carries its source. None of it replaces your allergist.

The short answer: the category is not all-or-nothing

If you read one section, read this. Legumes do not behave as a single block. What you do depends on which group a legume is in.

  • Most legumes are over-avoided, not cross-reactive. A peanut allergy does not predict soy, peas, lentils, or chickpeas. Cross-sensitization on testing is common, but clinical reactions to the other legumes are uncommon, so the default is do not blanket-avoid the whole aisle. Whether your child can eat a specific legume is still an allergist decision, not a blanket green light.
  • Lupin is the real exception, and fenugreek is the second. Lupin genuinely cross-reacts with peanut, and lupin flour hides in many breads and baked goods. Fenugreek, a legume used as a spice, also cross-reacts with peanut and usually arises from a primary peanut allergy. Do not assume a peanut-allergic person tolerates either.
  • The within-legume triad: lentil, chickpea, and pea travel together. These three show genuine clinical cross-reactivity with one another, so an allergy to one is a real reason to test the other two.
  • The shared mechanism, and the red flag. The serious reactions run through one shared family of seed-storage proteins, and a positive test to a storage protein is a red flag, not a reassuring finding. The protein-by-protein detail lives on its own page, the seed storage protein cluster.

The rule that ties it together: the legume aisle is not all-or-nothing. The reassuring default is do not over-avoid, with three honest exceptions (lupin, fenugreek, and the lentil-chickpea-pea triad). A positive panel is a reason to test, not a verdict to drop every legume, and direction matters: which legume is the original allergy changes the question, so test each direction with your allergist.

The shared mechanism: seed storage proteins, named here and explained on the syndrome page

For a family page the mechanism comes first, because it is what makes the map make sense, and then it hands off. Most serious legume reactions, and most real cross-reactions between two legumes, run through one shared family of proteins: the seed storage proteins, the 2S albumins, the 7S vicilins, and the 11S legumins.

Two practical points come out of that. First, these proteins are built to survive: they are heat-stable and resist digestion, so cooking does not defuse them. Thoroughly cooked lentils can still trigger a reaction. Second, and most important, a positive test to a legume storage protein is a red flag for whole-body reactions, not a reassuring low-risk result. It is a reason for strict avoidance and an action plan, confirmed with your allergist, never a reason to relax.

This is also why a positive legume panel is not the verdict it looks like. Storage proteins across legumes are related enough that a test can light up for several legumes at once, but the immune system reacting in a tube is not the same as the body reacting on the plate. Where the proteins are close enough that the test and the real reaction track together is exactly the lupin-peanut link, the fenugreek-peanut link, and the lentil-chickpea-pea triad below. Where they are not, which is most of the aisle, the positive panel is the over-avoidance trap, not a reaction.

This page names the mechanism and stops there on purpose. The protein-by-protein deep dive, how the storage proteins differ across peanut, the tree nuts, the seeds, and the legumes, and which homologies drive which reactions, is its own page: the seed storage protein cluster. Soy has a second, separate cross-reactivity story through a birch-pollen protein, Gly m 4, which usually drives a milder oral-allergy pattern rather than the systemic storage-protein one; that belongs to the oral allergy syndrome and birch cross-reactivity page. If you want the “why” at the protein level, those are where it lives. What this page does next is turn the mechanism into the food map.

The member foods, grouped by how strongly they cross-react

These are the legumes, grouped by what the cross-reaction actually does to the plate. Each group carries the verified record behind it.

The real cross-reactor: lupin with peanut

Lupin is the legume that genuinely cross-reacts with peanut. Lupin allergy can develop in a peanut-allergic person through cross-reactivity between their proteins, so a peanut-allergic person should not be assumed to tolerate lupin. Lupin is also the legume people least recognize, because lupin flour is a common wheat-flour substitute that hides in breads and baked goods (see the hidden-sources section). This is the one within-legume edge that, on a peanut allergy alone, is worth actively flagging to your allergist.

The second cross-reactor: fenugreek with peanut

Fenugreek is a legume used as a spice (in curry blends, some teas and supplements), and it cross-reacts with peanut. Fenugreek allergy usually arises from a primary peanut allergy rather than the other way round. As with lupin, do not assume a peanut-allergic person tolerates fenugreek, and be aware it travels under a spice name rather than a “legume” label.

The within-legume triad: lentil, chickpea, and pea

Lentil, chickpea, and pea show genuine clinical cross-reactivity with one another, not just shared positive tests. Lentil and pea co-react frequently, chickpea and lentil are among the most clinically cross-reactive legume pairs, and chickpea and pea co-react as well. In practice an allergy to one of these three is a real reason to test the other two, especially in children, and especially in Mediterranean and South Asian diets where these legumes are dietary staples. This is the within-family exception to the reassuring default: inside this triad, do not assume tolerance, test.

Where the reassurance is also the exception: lupin and fenugreek with the rest of the legumes

People understandably ask whether lupin or fenugreek is safe for a child already allergic to lentil, chickpea, or pea. The honest answer on our verified floor is that this is not established as safe, so the page does not offer a reassurance there. Lupin in particular is one of the legumes most likely to cause a genuine reaction in legume-allergic people, so a child allergic to lentil, chickpea, or pea should not be assumed to tolerate lupin, and the direction with fenugreek is unsettled enough that it is an each-food, allergist-guided question rather than an assumption in either direction.

Do not over-avoid: peanut with soy, peas, lentils, and chickpeas

This is the part most often gotten wrong, and it is gotten wrong in the over-cautious direction. Beyond lupin and fenugreek, a peanut allergy does not predict the other legumes. Most people with a peanut allergy tolerate soy, peas, lentils, and chickpeas, because clinical cross-reactivity is low even though cross-sensitization on testing is common. That is the cleared correction, and it is real. What it is not is a per-food green light: whether your specific child can eat soy, or chickpea, or pea, is still confirmed with your allergist before introducing, because the correction is about not blanket-avoiding the aisle, not about clearing one food at the kitchen table. And the correction runs in the peanut-outbound direction. A pea, lentil, or chickpea allergy is a different question (the triad above), and direction matters enough that “I am allergic to pea, am I allergic to peanut” is not the same question as “I am allergic to peanut, am I allergic to pea,” and is one to test with your allergist rather than read off this page.

Where the legumes hide

Because a single legume allergy can put a real second legume on the avoid list (lupin or fenugreek on a peanut allergy, the other two on a triad allergy), and because most of these legumes are not US major allergens, the hidden-source net for this page covers the whole group. The recognition problem is the point: the food is on the plate under a name you would not connect to “legume.” This is family territory, because it is mostly a labeling story.

Lupin, the hidden cross-reactor, in bread and baked goods. Lupin flour is a common wheat-flour substitute in European and imported breads, croissants, brioche, pasta, pizza dough, vegan meat substitutes, and gluten-free bakery. In the EU and UK lupin is a mandatory declarable allergen shown in bold, and it is mandatory in Australia and New Zealand, but in the US it is not a FALCPA major allergen, so US labels may not flag it. It stays hidden mostly because many people do not recognize the word, so read ingredient lists closely, especially on imported or gluten-free baked goods.

Chickpea, in hummus and falafel, and under “besan” and “gram flour.” Hummus, falafel, and channa are chickpea-based and ubiquitous in Middle Eastern, Mediterranean, and South Asian cuisine, and chickpea allergy can cause anaphylaxis. Papadum is often chickpea (gram) flour but is sometimes lentil-based, so check. The bigger trap is the flour: besan, also labeled gram flour, chickpea flour, or garbanzo flour, is ground chickpea, central to South Asian cooking (pakora, bhaji, dhokla) and increasingly common as a naturally gluten-free flour (socca, farinata, gluten-free breads and crackers). Treat all of those names as chickpea, and read gluten-free labels closely, since “gram flour” is not obviously chickpea.

Lentil, in dal and mixed dishes, and not defused by cooking. Lentil is a common pediatric legume allergen, especially in Mediterranean and South or East Asian populations, and can cause anaphylaxis. It is the base of dal and dhal and many soups and stews, and its allergens are heat-stable, so thoroughly cooked lentils can still trigger a reaction. In mixed dishes lentils are often unlabeled, so confirm with the cook.

Pea, the newest hidden source, as “pea protein.” Pea protein isolate is rapidly replacing soy as a soy-free, dairy-free protein in plant-based meats, dairy substitutes, protein bars, shakes, and some products, often labeled only “pea protein,” “pea fibre,” or “vegetable protein.” Pea is not a US major allergen, so it is easy to miss, and it is a real exposure for anyone already allergic to pea. If you switch to pea-based products, read labels closely.

Soy, in concentrated protein forms, not just the obvious ones. Soy is a US and EU major allergen, so it is usually declared, but the higher-risk forms are easy to underrate. Textured soy or vegetable protein (TVP, TSP) and hydrolyzed soy protein are concentrated soy-protein ingredients, not the trace-protein class like lecithin, so they are a genuine higher-risk source for a soy-allergic reader. They are common in vegan and meat-substitute products and in processed and deli meats. Watch the generic term “hydrolyzed vegetable protein,” which can be soy, and which US labels must declare the soy source for.

Fenugreek, under a spice name. Fenugreek travels in curry and spice blends, some teas, and some supplements, so it does not announce itself as a legume. If fenugreek is on the avoid list because of a peanut or fenugreek allergy, scan spice blends and supplement labels by name.

What is NOT cross-reactive, and where over-avoidance creeps in

The cleared reassurance on this page is the big one, and it is what makes the whole page net-reassuring. What stays held is any blanket “go ahead and feed it” claim about a specific legume; that decision stays with your allergist.

A peanut allergy does not mean avoiding all legumes. This is the correction worth repeating, because it is the most common and most costly over-avoidance in the whole category. Most people with a peanut allergy tolerate soy, peas, lentils, and chickpeas; cross-sensitization on a test is common, but clinical reactions to the other legumes are uncommon. Cutting the entire legume aisle on the strength of a peanut allergy, or a positive legume panel, costs diet, nutrition, and stress and usually buys no safety. The correction is do not blanket-avoid; the per-food decision (can my child eat soy, can my child eat chickpea) is still confirmed with your allergist, never read off as an automatic yes here.

Most legumes are the over-avoided ones. Soy, peas, lentils, and chickpeas are, on a peanut allergy alone, the over-avoided foods, not the cross-reactive ones. They are tested with your allergist, not dropped on a positive panel. The genuine within-legume exceptions are the three that travel together (lentil, chickpea, and pea, among themselves) and the two real peanut cross-reactors (lupin and fenugreek); everything outside those is where over-avoidance creeps in.

A blanket per-food “you can feed it” is not cleared, and it stays that way here. It is tempting to convert the reassuring default into a yes for one specific food at the table. The evidence supports do-not-blanket-avoid, not a per-food clearance, so the honest instruction is the same as in the food map: do not over-avoid, and let your allergist decide each food and each direction. Where a child has never been tested, “we avoid it to be safe for now” is a defensible holding pattern to revisit with testing, not a permanent verdict.

Where studies disagree

One area is genuinely unsettled, and the honest move is to publish the gap rather than pick a side. This is also what makes the “do not over-avoid” line trustworthy instead of glib.

Co-sensitization versus real-world reactivity. People with one legume allergy very frequently test positive to several others, because legume storage proteins are related enough that a panel co-fires across the group. When those same people are challenged under supervision, real reactions to the other legumes are much less common than the testing predicted, with the peanut-to-other-legume direction being the clearest example of a large gap. Both findings are true: the panel genuinely co-fires, and the clinical reactions genuinely do not follow at the same rate. The studies are measuring two different things, sensitization and reaction, and the gap between them is the whole reason a positive panel is a reason to test, not a verdict to avoid every legume. The lentil-chickpea-pea triad is the exception that proves the rule: there the co-firing and the real reactions track closely, because the proteins are close enough. Lupin sits with the triad as a real cross-reactor; peanut-to-soy sits at the over-avoided end, where the test fires far more often than a reaction occurs.

The test that answers the cross-reactivity question for legumes is not the broad panel; it is component-resolved diagnostics and, where the answer is still unclear, a supervised challenge.

A standard skin-prick or whole-extract blood test, and especially a broad legume panel, tells you mostly that the immune system has noticed the legume group and that the proteins co-fire, which they usually do. Component testing breaks the result down protein by protein, which is what separates the storage-protein pattern (the serious, cross-reactivity-meaningful kind) from the milder birch-pollen pattern that drives soy’s oral-allergy reactions. For the cross-reactivity question specifically, a positive result on the shared storage proteins is the signal that a real edge (lupin-peanut, fenugreek-peanut, or the triad) may apply, and it is a red flag rather than a reassurance, as the mechanism section explains.

When component testing and history still disagree, or when you want to know whether your child can actually eat a specific legume, the supervised oral food challenge is the reference standard. It is the one test that distinguishes a flag from a fight for a specific food, and it is done with your allergist, never at home, and never for a cross-reactive food on a hunch. It is also where the reassuring default becomes a specific yes for your child: the over-avoidance correction says do not blanket-avoid, and the challenge is how a specific legume gets cleared for a specific person.

How to act on this

The whole category reduces to a few moves:

  • Do not over-avoid: a peanut allergy does not mean avoiding soy, peas, lentils, and chickpeas. A positive legume panel is a reason to test, not a list of foods to fear. The per-food yes is an allergist decision, not a home one.
  • Flag the two real peanut cross-reactors: lupin (hidden as lupin flour) and fenugreek (hidden as a spice). Do not assume a peanut-allergic person tolerates either.
  • Test the triad: an allergy to lentil, chickpea, or pea is a real reason to test the other two with your allergist.
  • Mind direction: which legume is the original allergy changes the question. Pea-to-peanut is not the same as peanut-to-pea; test each direction rather than assume.
  • Learn the hidden names: lupin flour, besan and gram flour, dal, pea protein, TVP and hydrolyzed soy protein, fenugreek in spice blends. Most of these legumes are not US major allergens.

Cross-reactivity is the part of a legume allergy most easily turned into avoiding a whole food group, and also the part where a clear map saves the most worry. The map is mostly reassuring, with three honest exceptions and one shared protein rule.

Frequently asked questions

If I am allergic to peanut, do I have to avoid all legumes?

No. A peanut allergy does not mean avoiding soy, peas, lentils, and chickpeas. Cross-sensitization on a test is common, but clinical reactions to the other legumes are uncommon, so the default is do not blanket-avoid the legume aisle. The real exceptions are lupin and fenugreek. Whether your child can eat a specific legume is confirmed with your allergist before introducing it, not read off automatically.

Which legumes actually cross-react with peanut?

Lupin and fenugreek. Lupin genuinely cross-reacts with peanut and hides as lupin flour in breads and baked goods. Fenugreek, a legume used as a spice, cross-reacts with peanut and usually arises from a primary peanut allergy. Soy, peas, lentils, and chickpeas are the over-avoided ones, not the cross-reactive ones, on a peanut allergy alone.

My child is allergic to lentil. Should we test chickpea and pea?

Yes, this is a real reason to test. Lentil, chickpea, and pea show genuine clinical cross-reactivity with one another, not just shared positive tests. An allergy to one of these three is a reason to test the other two with your allergist, rather than assume either tolerance or reaction.

Is a positive legume panel the same as a legume allergy?

No. A positive test to several legumes at once is common, because legume proteins are related enough that the panel co-fires, but the immune system reacting on a test is not the same as the body reacting on the plate. A positive panel is a reason to test the specific foods with your allergist, not a verdict to avoid every legume. Direction matters too: which legume is the original allergy changes the question.

Does a positive test to a legume storage protein mean it is a mild allergy?

No. A positive result on a legume seed-storage protein (the 2S albumins, 7S vicilins, and 11S legumins) is a red flag for whole-body reactions, not a reassuring or minor finding, and these proteins are heat-stable, so cooking does not destroy them. The protein-level “why” is on the seed storage protein cluster page.

References and medical review

This page is pending independent medical review; the note at the top of the page applies until a reviewer is assigned. The verified cross-reactivity, hidden-source, and reassurance claims resolve to the project’s conservative cross-reactivity floor, each carrying its own tier-1 source there. The component biology (the seed-storage-protein families, the heat-stable and digestion-resistant behavior, soy’s separate Gly m 4 birch-pollen track, and the broad-panel over-diagnosis gap) resolves to the consolidated legume research reports still pending final review. Figures not yet pinned to a stable source, such as the within-triad challenge percentages and the peanut-to-other-legume reaction rate, are omitted rather than stated. The held per-food introduction green lights are not rendered as clearances, and the pea-to-peanut edge is not established on the verified floor, so this page asserts it in neither direction.

← Cross-reactivity guides