Why Brussels Sprouts Stopped Tasting Bitter

If you are old enough to remember Brussels sprouts as the worst thing on the plate, and you have tried them recently and been surprised, you are not misremembering and you have not simply grown up. The sprouts changed. Plant breeders worked out which specific chemicals made them bitter, built a laboratory test that could measure those chemicals in a leaf sample, showed the traits were inherited in a way selection could act on, and bred the bitterness down. This page tells that story with the actual papers attached, separates the well-documented parts from the parts that only exist in newspaper retellings, and then explains the other half of the answer — that people genuinely differ, at the level of a single gene, in how bitter the same sprout tastes.

Table of Contents

  1. The Change People Noticed Is Real
  2. What Actually Makes a Sprout Bitter
  3. The Dutch Breeding Programme
  4. What Is Documented and What Is Retold
  5. Conventional Breeding, Not Genetic Modification
  6. Why the Same Sprout Tastes Different to Different People
  7. Supertasters, Non-Tasters and What the Label Misses
  8. The Trade-Off: Did We Breed Out the Medicine?
  9. How Much of the Improvement Is Just Better Cooking?
  10. Making Sprouts Taste Good on Purpose
  11. Key Research Papers
  12. Connections
  13. Featured Videos

The Change People Noticed Is Real

Brussels sprouts spent most of the twentieth century as a punchline. Surveys of least-liked vegetables put them at or near the bottom in Britain and the United States for decades, and the dislike was strong enough to be a running joke rather than a preference. Then, over roughly the 1990s and 2000s, something shifted. Restaurants started putting them on menus deliberately rather than as an obligation. Home cooks who had avoided them for thirty years tried them again and found they were eating a different vegetable.

Two explanations get offered for this. The first is that the cooking changed — that people stopped boiling sprouts into grey submission and started roasting them, and that roasting would have made the old sprouts taste good too. The second is that the sprouts themselves changed. Both are partly true, and this page argues that the second is the more surprising and better-evidenced half of the answer, because there is a chain of published work behind it that names the compounds, measures them, and shows they can be selected.

The reason it matters beyond dinner is that it is one of the clearest examples anywhere of deliberately breeding a flavour defect out of a food crop using ordinary genetics. Most crop breeding in the same period chased yield, disease resistance and shelf life, and flavour was what got sacrificed to them — the modern supermarket tomato is the standard cautionary tale. Brussels sprouts went the other way.

Back to Table of Contents

What Actually Makes a Sprout Bitter

Brussels sprouts, like every other Brassica oleracea vegetable, carry a family of sulfur-containing compounds called glucosinolates. In the intact plant they are chemically inert and stored separately from the enzyme that activates them. Damage the tissue — chewing, chopping, an insect bite — and the enzyme myrosinase meets its substrate and cuts the glucosinolate apart, producing sharp, reactive breakdown products. That is the plant's chemical defence, and it is also the source of the flavour. The shared chemistry is covered in depth on the cabbage glucosinolates page, which applies to the whole family; what follows is what is specific to sprouts.

Sprouts are near the top of the family for total glucosinolate content. When Kushad and colleagues measured glucosinolates across the Brassica oleracea vegetable crops, Brussels sprouts came out as one of the richest, and their profile is dominated by two compounds that matter for taste:

  1. Sinigrin — the main aliphatic glucosinolate in sprouts. Its breakdown product is allyl isothiocyanate, the same compound that gives mustard, horseradish and wasabi their pungency. In small amounts it is sharp and appetising; in larger amounts it is harsh.
  2. Progoitrin — which breaks down to goitrin (chemically, 5-vinyloxazolidine-2-thione). This one is straightforwardly bitter, and it is also the compound behind the sprout's reputation as a goitrogen, discussed on the cooking and tolerance page.

The identification of goitrin as the bitter principle of cooked sprouts was made by Fenwick and Griffiths in 1981, working at what was then the Food Research Institute in Norwich. Two years later Fenwick, Griffiths and Heaney published a fuller account of bitterness in Brussels sprouts and its relationship to glucosinolates and their breakdown products. So by the mid-1980s the chemistry was on the table: a named, measurable compound was responsible for a large part of the problem.

What that work did not do was translate into different sprouts in shops, because knowing which compound to remove is only the first of four things a breeder needs. You also need a way to measure it quickly in thousands of plants, evidence that the trait is heritable rather than driven by soil and weather, and a source of low-bitterness genetics that has not already been discarded. All four arrived in the 1990s, and they arrived in the Netherlands.

Back to Table of Contents

The Dutch Breeding Programme

The published core of the story belongs to Hans E. van Doorn and colleagues at the Department of Biotechnology, Novartis Seeds BV in Enkhuizen, a seed-breeding town in North Holland. Three papers and a doctoral thesis, all between 1998 and 1999, lay out the work in order.

  1. Measure it fast. In early 1998, in the Journal of Agricultural and Food Chemistry, the group described antibody-based ELISA assays that could quantify sinigrin and progoitrin specifically in Brussels sprouts. This is the unglamorous step that makes plant breeding possible at all. Full glucosinolate profiling by chromatography is accurate and far too slow to screen a breeding population; an immunoassay you can run in plates turns a research measurement into a selection tool.
  2. Prove the compounds are the problem. Later in 1998, in the Journal of the Science of Food and Agriculture, the group published the paper the whole story rests on: The glucosinolates sinigrin and progoitrin are important determinants for taste preference and bitterness of Brussels sprouts. Chemical measurements were tied to sensory panel judgements, so the link between the two named compounds and what people actually reported tasting was established rather than assumed.
  3. Show it can be inherited. In 1999, in Euphytica — the European plant-breeding journal — the group published a quantitative-genetic analysis of progoitrin and sinigrin content in Brussels sprouts. Quantitative inheritance means the trait is controlled by several genes with additive effects, which is exactly the kind of trait that responds steadily to selection generation after generation.
  4. Write it up as a programme. Van Doorn's Wageningen University thesis, titled Development of vegetables with improved consumer quality: a case study in Brussels sprouts, presents the whole thing as one project rather than four papers, and it is the single best source if you want the reasoning rather than the results.

Read together, the sequence is unusually clean. A flavour defect was reduced to two named molecules; a fast assay was built for them; the assay was validated against human tasters; the trait was shown to be heritable; and the work was carried out inside a commercial breeding company whose business was selling seed to growers. That last point is the one that turns a paper into a vegetable in a shop.

Back to Table of Contents

What Is Documented and What Is Retold

The popular version of this story, which has been retold in newspapers, on radio and across food writing since the late 2010s, adds a specific final step: that breeders went back to old, low-bitterness sprout varieties that had been abandoned for poor yield, crossed them with modern high-yielding lines, and selected the offspring that kept the yield and lost the bitterness. It is a satisfying ending and it is entirely plausible — it is, in fact, standard practice in plant breeding, and it is what the quantitative-inheritance result implies you would do next.

It is worth being straight about where the evidence stops. The peer-reviewed record establishes the compounds, the assay, the sensory link and the heritability. The specific narrative of crossing back to named heritage lines, and the claim that this is what put better-tasting sprouts on shelves in a particular year, comes from interviews and trade reporting rather than from a paper you can read. Commercial breeding programmes are not usually published in detail, for obvious reasons, so the absence of a citation is not evidence against it — but it is not the same kind of claim as the four above, and this site does not pretend otherwise.

Two further things are documented and often left out of the retelling. First, this was not the only thing changing: Brussels sprouts had already moved from open-pollinated varieties to F1 hybrids over the 1960s and 1970s, which transformed uniformity and made once-over mechanical harvest possible, and hybrid breeding gives a breeder much tighter control over which traits travel together. That history is on the Brussels sprouts history page. Second, glucosinolate content varies substantially with growing conditions, plant age, position on the stem and time of harvest, so any given bag of sprouts sits somewhere in a range, and the variety is only part of what determines where.

Back to Table of Contents

Conventional Breeding, Not Genetic Modification

People sometimes assume that a deliberate, targeted change in a plant's chemistry must involve genetic engineering. It did not here. Everything described above is conventional crossing and selection: pollinate, grow the offspring, measure them, keep the best, repeat. The ELISA assay is a laboratory tool used to decide which plants to keep, not a way of altering them. No gene was inserted, edited or moved between species.

That distinction matters for anyone who avoids genetically modified food. It also matters for a different reason: it means the change was slow and incremental. Selection on a quantitative trait moves the population average a little each generation. There was never a moment when bitter sprouts stopped and mild sprouts started; there was a gradual shift across a couple of decades, which is exactly what people describe experiencing.

Back to Table of Contents

Why the Same Sprout Tastes Different to Different People

The other half of the answer has nothing to do with the vegetable. If you have ever watched one person eat sprouts happily while another gags at the same dish and concluded that one of them is being difficult, the genetics say otherwise.

The story starts with an accident in a chemistry laboratory in 1931, when the compound phenylthiocarbamide (PTC) escaped as dust and some of the people in the room tasted intense bitterness while others tasted nothing at all. PTC tasting became one of the classic teaching examples in human genetics, and for seventy years nobody knew which gene it was. In 2003, Kim, Jorgenson, Coon and colleagues reported in Science that they had positionally cloned it: the locus is TAS2R38, a bitter taste receptor gene on chromosome 7. Three coding variants define two common haplotypes, usually written PAV (the taster form) and AVI (the non-taster form).

Two years later Bufe, Breslin and colleagues showed in Current Biology that the receptor variants behave differently in the laboratory in exactly the way the psychophysics predicts — the PAV receptor responds to PTC and to the related compound propylthiouracil (PROP) at far lower concentrations than the AVI receptor. That closed the loop from gene to protein to perception.

The step that connects all of this to dinner came in 2006, when Sandell and Breslin published, again in Current Biology, a paper with a title that says it plainly: Variability in a taste-receptor gene determines whether we taste toxins in food. They tested people of known TAS2R38 genotype on actual Brassica vegetables and found that PAV/PAV individuals rated them substantially more bitter than AVI/AVI individuals. The receptor that responds to PTC and PROP also responds to the naturally occurring compounds in these vegetables — which makes chemical sense, because PTC, PROP and goitrin all carry the same reactive thiourea-like grouping, an N–C=S motif, that the receptor is tuned to.

So the honest summary is this: “I hate Brussels sprouts” can be a genotype. A PAV/PAV taster eating a high-progoitrin sprout is having a genuinely more intense bitter experience than an AVI/AVI person eating the same forkful, and no amount of encouragement changes the receptor.

There is one more twist worth knowing. Wooding and colleagues examined the pattern of variation at the PTC locus across human populations in 2004 and found signatures inconsistent with simple neutral drift — a pattern suggesting that both the taster and non-taster forms have been maintained rather than one being replaced. Whatever the details, the sensible reading is that being able to detect these compounds and being able to ignore them were each useful under different circumstances, which is a far more interesting explanation than one of them being a defect.

Back to Table of Contents

Supertasters, Non-Tasters and What the Label Misses

The word supertaster gets attached to this research, and it is worth unpicking because it quietly merges two different things.

TAS2R38 genotype is a specific, binary-ish thing: which version of one bitter receptor you carry. It predicts a great deal of the variation in how bitter PTC, PROP and thiourea-like compounds — including goitrin — taste to you. General taste intensity is a different and broader phenomenon, associated with how densely packed the fungiform papillae on the tongue are and with overall oral sensation, and it affects sweetness, saltiness, fat texture and burn as well as bitterness. People who are high on both tend to get called supertasters, but the two are separable, and a person can carry the taster genotype without being unusually sensitive to everything else.

Two further caveats keep this honest. TAS2R38 is one of roughly two dozen bitter receptor genes in humans, and the glucosinolate breakdown products in a sprout are not a single compound, so genotype at one locus explains a lot but never all of the variation. And self-reported vegetable liking is influenced by culture, childhood exposure and how the vegetable was cooked, which is why population studies find real but modest associations. A 2023 analysis of United States national survey data by Ma and Lu, for instance, examined bitter taste sensitivity alongside cruciferous vegetable intake and metabolic outcomes and found relationships that are statistically visible without being large enough to determine any individual's diet.

Back to Table of Contents

The Trade-Off: Did We Breed Out the Medicine?

This is the uncomfortable question, and it deserves a direct answer rather than reassurance.

Glucosinolates are not only the bitter compounds. They are also the precursors of the isothiocyanates that carry most of the health claims made for cruciferous vegetables — the compounds that trigger the cell's own antioxidant and detoxification machinery, discussed on the detox enzymes page. If you breed glucosinolates down to make a sprout taste better, you are lowering the very thing the health argument rests on. That is a real tension, and anyone who tells you otherwise has not thought about it.

Three things make it less bad than it first sounds.

  1. The two target compounds are not equally valuable. Progoitrin's breakdown product, goitrin, is bitter and is the compound with genuine antithyroid activity. Reducing progoitrin lowers bitterness and lowers the one glucosinolate-derived compound with a documented downside, which is close to a free win. Sinigrin is the harder case: allyl isothiocyanate is one of the bioactive isothiocyanates, so lowering sinigrin does cost something.
  2. The indole pool is largely separate. Glucobrassicin and its relatives, which yield indole-3-carbinol and DIM, are a different chemical family from the aliphatic glucosinolates and are not the main drivers of the bitterness the breeders targeted. Selection against sinigrin and progoitrin does not automatically strip the indoles.
  3. The vegetable people eat beats the vegetable they avoid. A moderately lower glucosinolate content in sprouts that get eaten weekly delivers more isothiocyanate than a maximal content in sprouts that get pushed to the side of the plate. This is not a rhetorical dodge — every human trial in this field feeds people substantial quantities of sprouts, and compliance is the binding constraint in real life.

What nobody can currently tell you is the size of the trade-off, because there is no long-running published series measuring glucosinolate content in commercial Brussels sprout varieties by year. The honest position is that some reduction almost certainly happened, that it was concentrated in the compounds with the worst taste-to-benefit ratio, and that the magnitude is unquantified. If you want to be sure of a high isothiocyanate intake, eat a wider range of the family — watercress, arugula, radish and turnip have not been bred for mildness at all, and they taste like it.

Back to Table of Contents

How Much of the Improvement Is Just Better Cooking?

A substantial share, and pretending otherwise would be as one-sided as ignoring the breeding.

The mid-century default was boiling, often for fifteen minutes or more, sometimes with the base cut in a cross “so they cook evenly”. Prolonged boiling does three bad things at once. It leaches water-soluble compounds into the water, including vitamin C and a large fraction of the glucosinolates. It drives off and rearranges the sulfur chemistry into the volatile compounds responsible for the smell that fills a house. And it collapses the texture, which removes the contrast that makes the flavour tolerable. The measured losses are covered on the cooking page; the sensory result is the grey sprout of memory.

Roasting does the opposite. There is no cooking water, so nothing leaches. High dry heat caramelises the sprout's own sugars and drives Maillard browning, producing sweet and savoury compounds that were not there before. And sweetness genuinely suppresses bitterness — Beck and colleagues demonstrated exactly this masking effect for bitter compounds in Brassica vegetables using sucrose, which is the formal version of what every cook who finishes sprouts with a little balsamic or maple already knows.

So the fair verdict is that both things happened, and they compound. Better varieties made the raw material less punishing; better technique stopped destroying it. Either alone would have helped; together they changed the vegetable's reputation.

Back to Table of Contents

Making Sprouts Taste Good on Purpose

If you know why they taste bitter, the fixes stop being folklore.

  1. Roast, do not boil. Halve them, toss in olive oil, cut side down on a hot tray, high oven, until the cut faces are properly browned and the outer leaves have gone crisp and dark. Browning is the flavour; timid roasting gives you a steamed sprout in a hot oven.
  2. Buy small ones. Small, tight sprouts are younger and generally milder. Large, loose, yellowing sprouts have had longer to accumulate and are the ones that taste like the reputation.
  3. Add a little sweetness and a little acid. Sweetness suppresses bitterness, which is the Beck result; acid brightens the whole dish. A splash of balsamic or lemon in the last minutes does both.
  4. Use fat. Olive oil, butter or the fat rendered from bacon or pancetta carries the aroma compounds and blunts harshness. This is a real physical effect, not indulgence.
  5. Cook them less than you think. Overcooking is what generates the sulfurous volatiles. Aim for tender-crisp at the core with a browned exterior.
  6. Try them raw and shredded. Finely shaved raw sprouts in a salad with lemon, oil, toasted nuts and something salty is a completely different food from a cooked sprout, and it converts a surprising number of sceptics.
  7. If you are a strong taster, use the tools rather than fighting it. Salt suppresses bitterness. Fat suppresses bitterness. Sweetness suppresses bitterness. Browning replaces it with something else. Stack all four.
  8. Repetition works, particularly for children. Repeated tasting of a bitter vegetable increases intake and liking over time; Mohd Nor and colleagues showed this for turnip in children and looked at whether bitter-taste sensitivity changed the result. The practical advice — small portions, offered often, without pressure — is well supported, and it is more effective than a single heroic serving.

Back to Table of Contents

Key Research Papers

  1. van Doorn HE, van der Kruk GC, van Holst GJ, Raaijmakers-Ruijs NCME, Postma E, Groeneweg B, Jongen WHF. The glucosinolates sinigrin and progoitrin are important determinants for taste preference and bitterness of Brussels sprouts. Journal of the Science of Food and Agriculture. 1998;78(1):30-38. — doi:10.1002/(SICI)1097-0010(199809)78:1<30::AID-JSFA79>3.0.CO;2-N
  2. van Doorn HE, van Holst GJ, van der Kruk GC, Raaijmakers-Ruijs NCME, Postma E. Quantitative determination of the glucosinolates sinigrin and progoitrin by specific antibody ELISA assays in Brussels sprouts. Journal of Agricultural and Food Chemistry. 1998;46(2):793-800. — doi:10.1021/jf970523z
  3. van Doorn JE, van der Kruk GC, van Holst GJ, Schoofs M, Broer JB, de Nijs JJM. Quantitative inheritance of the progoitrin and sinigrin content in Brussels sprouts. Euphytica. 1999;108(1):41-52. — doi:10.1023/a:1003600227319
  4. van Doorn JE. Development of vegetables with improved consumer quality: a case study in Brussels sprouts. Doctoral thesis, Wageningen University. — doi:10.18174/195752
  5. Fenwick GR, Griffiths NM. The identification of the goitrogen, (−)5-vinyloxazolidine-2-thione (goitrin), as a bitter principle of cooked Brussels sprouts. Zeitschrift für Lebensmittel-Untersuchung und -Forschung. 1981;172(2):90-92. — doi:10.1007/bf01042410
  6. Fenwick GR, Griffiths NM, Heaney RK. Bitterness in Brussels sprouts (Brassica oleracea L. var. gemmifera): the role of glucosinolates and their breakdown products. Journal of the Science of Food and Agriculture. 1983;34(1):73-80. — doi:10.1002/jsfa.2740340111
  7. Kim UK, Jorgenson E, Coon H, Leppert M, Risch N, Drayna D. Positional cloning of the human quantitative trait locus underlying taste sensitivity to phenylthiocarbamide. Science. 2003;299(5610):1221-1225. — doi:10.1126/science.1080190
  8. Bufe B, Breslin PAS, Kuhn C, Reed DR, Tharp CD, Slack JP, Kim UK, Drayna D, Meyerhof W. The molecular basis of individual differences in phenylthiocarbamide and propylthiouracil bitterness perception. Current Biology. 2005;15(4):322-327. — doi:10.1016/j.cub.2005.01.047
  9. Sandell MA, Breslin PAS. Variability in a taste-receptor gene determines whether we taste toxins in food. Current Biology. 2006;16(18):R792-R794. — doi:10.1016/j.cub.2006.08.049
  10. Wooding S, Kim UK, Bamshad MJ, Larsen J, Jorde LB, Drayna D. Natural selection and molecular evolution in PTC, a bitter-taste receptor gene. The American Journal of Human Genetics. 2004;74(4):637-646. — doi:10.1086/383092
  11. Beck TK, Jensen S, Bjoern GK, Kidmose U. The masking effect of sucrose on perception of bitter compounds in Brassica vegetables. Journal of Sensory Studies. 2014;29(3):190-200. — doi:10.1111/joss.12094
  12. Kushad MM, Brown AF, Kurilich AC, Juvik JA, Klein BP, Wallig MA, Jeffery EH. Variation of glucosinolates in vegetable crops of Brassica oleracea. Journal of Agricultural and Food Chemistry. 1999;47(4):1541-1548. — doi:10.1021/jf980985s
  13. Mohd Nor ND, Houston-Price C, Harvey K, Methven L. The effects of taste sensitivity and repeated taste exposure on children's intake and liking of turnip (Brassica rapa subsp. rapa). Appetite. 2021;157:104991. — doi:10.1016/j.appet.2020.104991
  14. Ma S, Lu S. Bitter taste sensitivity, cruciferous vegetable intake, obesity, and diabetes in American adults: a cross-sectional study of NHANES data. Food & Function. 2023;14(20):9243-9252. — doi:10.1039/d3fo02175k
  15. Ockendon DJ. Distribution of self-incompatibility alleles and breeding structure of open-pollinated cultivars of Brussels sprouts. Heredity. 1974;33(2):159-171. — doi:10.1038/hdy.1974.84

Back to Table of Contents

Connections

Back to Table of Contents