Press Release
Rice Seed Grew 5.30 Per Cent While China's Hybrid Area Fell 25 Per Cent and India Sowed Hybrids on 6 Per Cent of 43 Million Hectares — and No Supplier in the World Holds More Than 4.3 Per Cent of It
Two thirds of the world’s rice hectares sow seed that is never bought. That is the market.
Rice is the only major seed market in the world where the buyer is frequently not the customer. Across South and Southeast Asia a large share of the area is sown with seed held back from the previous harvest or exchanged inside a village, and much of the rest is a public variety multiplied and distributed at cost recovery. The competitor to a commercial bag is not another company’s hybrid. It is a free variety with a known yield.
We size global rice seed at USD 6.35 billion in 2025, reaching USD 9.60 billion by 2033, a 5.30 per cent CAGR. That is 8.10 per cent of the USD 78.40 billion global seed market. Commercial seed volume runs from 3.20 to 3.72 million tonnes, or 1.90 per cent a year, and value per tonne of commercial seed from USD 1,984 to USD 2,580, or 3.34 per cent.
The identity closes: 1.019 × 1.0334 = 1.0530. And that arithmetic corrects the way this market usually gets described. The gap between value growth and volume growth is not conversion. It is price and mix. Conversion is already inside the 1.90 per cent, because conversion is what makes commercial tonnage grow faster than planted area. Split the USD 3.25 billion of increase on an endpoint bridge and it divides into USD 1,031.9 million from volume, USD 1,908.1 million from value per tonne, and USD 310.1 million of interaction, being 31.7, 58.7 and 9.5 per cent. A log decomposition of the same endpoints gives 36.4 and 63.6, which is the 36/64 the forecast model publishes. Both are defensible. They are not the same calculation, and this release says which is which.
Now take conversion out. Hold commercial volume at 3.20 million tonnes and let only price and mix work: 2033 lands at USD 8,258 million rather than 9,600. USD 1,342 million, or 41.3 per cent of the whole increment, exists only because more growers bought seed. The other 58.7 per cent arrives whether or not a single additional tonne moves through a commercial channel. That is the proportion to argue about, and it is a very different proposition from the one the sector makes about itself.
The forecast path
| Year | Value, USD billion | Growth |
|---|---|---|
| 2025 | 6.35 | — |
| 2026 | 6.62 | 4.3% |
| 2027 | 6.98 | 5.4% |
| 2028 | 7.36 | 5.4% |
| 2029 | 7.75 | 5.3% |
| 2030 | 8.16 | 5.3% |
| 2031 | 8.60 | 5.4% |
| 2032 | 9.07 | 5.5% |
| 2033 | 9.60 | 5.8% |
2026 is the soft year at 4.3 per cent, and it is soft for a reason that has nothing to do with seed. World milled rice production reached a record 545.65 million tonnes in 2025/26; India alone produced a record 154.02 million tonnes and closed the year holding 59 million tonnes of stocks, or 38.3 per cent of its own crop. Soft paddy is the worst possible backdrop for selling a more expensive bag. USDA then forecasts 2026/27 output 2.3 per cent lower at 533.08 million tonnes, which is the sequence the model assumes: the seed budget is squeezed before the paddy price is restored. On the 2026 base of 6.62 the remaining seven years compound at 5.45 per cent.
Where the growth actually sits, by seed class
| Seed class | 2025, USD bn | Share of 2025 value | CAGR to 2033 | 2033, USD bn | Share of the increase |
|---|---|---|---|---|---|
| Hybrid rice seed | 2.44 | 38.4% | 6.3% | 3.98 | 47.3% |
| Certified inbred (varietal) seed | 2.79 | 43.9% | 4.4% | 3.94 | 35.3% |
| Public-sector and subsidised seed | 0.72 | 11.3% | 3.6% | 0.96 | 7.2% |
| Premium quality and specialty seed | 0.40 | 6.4% | 7.1% | 0.69 | 9.0% |
| Total | 6.35 | 100% | 5.30% | 9.56 | 100% |
The largest class by value is not the largest class by growth. Certified inbred seed is 43.9 per cent of 2025 value and 35.3 per cent of the increase; hybrid seed is 38.4 and 47.3; premium and specialty seed, at 6.4 per cent of value, supplies 9.0 per cent of the growth. The four lines compound to USD 9.563 billion against the 9.60 control, 0.4 per cent light, which is rounding and is disclosed here rather than buried.
The class split also reconciles into the cluster. Our hybrid seed report puts hybrid rice germplasm at USD 2.44 billion, which is the hybrid line above. Everything else is varietal seed, most of it certified inbred multiplied by private and cooperative producers, plus the public-sector volume priced at cost recovery rather than at value.
And by how the crop is established, which is the split that matters
| Establishment method | 2025, USD bn | Share of 2025 value | CAGR to 2033 | 2033, USD bn | Share of the increase |
|---|---|---|---|---|---|
| Transplanted paddy | 4.22 | 66.5% | 4.3% | 5.91 | 49.5% |
| Dry direct-seeded | 1.57 | 24.7% | 8.1% | 2.93 | 39.8% |
| Water-seeded, aerobic and mechanised | 0.56 | 8.8% | 6.5% | 0.93 | 10.7% |
| Total | 6.35 | 100% | 5.30% | 9.76 | 100% |
Dry direct seeding is 24.7 per cent of value and supplies 39.8 per cent of the increase. This is the only part of the market where three things move in the same direction at once: seed rate per hectare, the share of seed that is purchased, and the traits a grower will pay for. Rice was transplanted by hand for the entire green revolution and increasingly is not.
One disclosure the table needs. Those three lines compound to USD 9.764 billion against the 9.60 control, 1.7 per cent over, where the class split came in 0.4 per cent under. The establishment rates average 5.43 per cent weighted against a 5.30 headline. That is a real inconsistency in a table the report calls the one to watch, and the release carries the rates and the gap rather than the total.
The mechanism is measured, not asserted. On IRRI’s 2025–26 winter–spring pilot in Dong Nai, mechanised dry direct seeding with fertiliser deep placement cut seed rates by about 25 per cent, or 20 kilograms per hectare, while reducing irrigation water use by more than 20 per cent, cutting greenhouse gas emissions by up to 30 per cent, and raising yields 15–20 per cent against conventional establishment. Broadcast direct seeding does the opposite to volume and typically uses two to three times the seed of a transplanted hybrid crop. Net across Asia, volume rises, because the shift out of transplanting is running faster than the shift into precision drilling.
What a direct-seeded field needs is anaerobic germination tolerance, early seedling vigour, herbicide tolerance, lodging resistance and a short duration that fits a two- or three-crop rotation. Those are establishment traits, not yield genes. The breeding programmes that hold them are the ones gaining share as labour leaves the transplanting systems of Asia.
China: a high share on a falling base
China commercialised hybrid rice in the 1970s and dominated the technology for four decades. It is now the only large rice market in the world where hybrid penetration runs backwards. The area under hybrid rice has fallen by roughly 5 million hectares, or 25 per cent, since 1995, as inbred varieties closed much of the yield gap, as labour left the Yangtze basin, and as consumers began paying for eating quality that the early hybrids did not deliver. The share remains above 60 per cent, the highest of any large market. Penetration and growth are not the same variable, and in China they now point in opposite directions.
Yuan Longping High-Tech’s 2025 accounts are the clearest evidence available that the Chinese hybrid rice business has stopped growing. Rice seed revenue fell 1.69 per cent to RMB 1.974 billion, or about USD 275 million, while gross margin improved 1.98 percentage points to 39.58 per cent. Group revenue was RMB 8,477 million. Group net profit rose 45.63 per cent on a Brazilian corn business that grew 16.51 per cent to RMB 3.661 billion, and research and development spending was RMB 763 million, or 9.00 per cent of revenue.
Set those lines beside each other and the arithmetic is blunt. Rice seed is 23.3 per cent of Longping’s revenue. Corn is 43.2 per cent, or 1.9 times as much. The hybrid rice specialist is now, financially, a corn company with a rice heritage. Its capital has followed: a RMB 1.2 billion private placement to CITIC Agriculture that lifted the controlling stake from 17.36 to 25.93 per cent, a RMB 1 billion external raise at its corn subsidiary Lianchuang Seeds, and disposals of a 25.38 per cent stake in Huazhi Biotech and 34 per cent of the vegetable seed unit Derit to Mitsui. Rice seed was not the recipient of new money in 2025. It was the stable cash generator that made the corn expansion financeable.
China is still USD 2.35 billion, or 37.0 per cent of world value, and it grows at 4.2 per cent, so it contributes 28.2 per cent of the entire increase on price and mix alone. It is the largest market in the world and the largest pool of breeding capability. It is not where the incremental tonnes come from, and modelling it as a growth market is the single most common error in this segment.
India: six per cent of 43 million hectares
India cultivates hybrid rice on about 6 per cent of its 43 million hectares, though the core eastern states of Bihar, Jharkhand, Chhattisgarh and Odisha run closer to 15 per cent. Bayer’s India management reported kharif 2025 hybrid acreage 6 to 8 per cent higher year on year on improved procurement prices and return on investment. Hybrids carry a 10–25 per cent yield advantage on Bayer’s estimate and 15–20 per cent on SeedWorks’.
Put a number on the headroom. Six per cent of 43 million hectares is 2.58 million. Fifteen per cent is 6.45 million. The gap is 3.87 million hectares, one and a half times the hybrid area India has today, and it is the largest conversion pool in this market by an order of magnitude. It is being won state by state and district by district, which is why SeedWorks can be ahead in Jharkhand, Chhattisgarh, Madhya Pradesh, Odisha, Gujarat and Nepal in a season when it is merely in line to slightly ahead overall because of Uttar Pradesh and Bihar.
India is USD 0.68 billion, 10.7 per cent of world value, growing at 7.4 per cent, and supplies 16.1 per cent of the increase. A third of world value growing at four and a tenth growing at seven is the entire regional story, and it is why Asia-Pacific at USD 4.95 billion grows at 5.45 per cent, in line with the global rate rather than above it.
The price ceiling is a free variety, and it shows in the margin
There is no tier-1 position in rice seed. The largest private supplier in the world earned about USD 275 million from rice seed in 2025, or 4.3 per cent of global value, in a group ranked seventh among world seed companies. We estimate the five largest private suppliers together hold under 15 per cent. In corn, two companies hold roughly 60 per cent, and Bayer alone about 32.
The reason is structural rather than commercial. Rice germplasm is a public good almost everywhere it is grown: IRRI and the national programmes release varieties, governments multiply them, and subsidy programmes distribute them. That caps both price and share, and it is why gross margins in this market sit in the high thirties rather than the fifties. Longping’s 39.58 per cent is not a weak result for a hybrid seed business. It is the ceiling.
The report’s conversion coefficient is worth testing against its own volume series. It states that every percentage point of rice area moving from farm-saved or public variety to a purchased bag adds roughly USD 60 million of value. The volume series says the same thing from the other end: 520 thousand tonnes of additional commercial seed over eight years, valued at 2025 prices, is USD 1,031.9 million, or USD 129 million a year. Spread across 1.90 points of annual volume growth that is USD 68 million a point. The published coefficient is the conservative version of the model’s own arithmetic, which is the right direction for it to be wrong in.
The bag is priced against the public variety, and the traits are priced against the labour market
Every strategic conclusion in this market follows from two facts. The first is that in most of Asia the alternative to a commercial bag is free seed with a known yield, so the premium has to be defended in the district where it is sold, on a measured return, every season. Where the state buys paddy at a floor, hybrid economics work. Where it does not, the seed premium has to be earned again from scratch.
The second is that labour scarcity is doing more to the seed rate than any breeding programme. As transplanting labour becomes unavailable, growers switch to direct seeding, and the same hectares, sown differently, need more seed. Volume therefore grows for a reason that has nothing to do with planted area. That is why commercial volume can rise 1.90 per cent a year while the crop does not expand at all.
None of this is a technology story in the way corn seed is. Corn grows because a hectare can absorb more trait value every year. Rice grows because a farmer stops saving seed. The distinction explains the rate: 5.30 per cent against corn’s 4.60, on two markets that are nothing alike. Rice moves 74 per cent of the tonnage corn does and earns 26 per cent of the money. A tonne of rice seed is worth USD 1,984 against corn’s USD 5,721, or 34.7 per cent. Measured per hectare of the crop, rice is the least monetised major seed market in the world, and at USD 6.35 billion it is the fourth-largest by value behind corn’s USD 24.60 billion, soybean’s USD 12.60 billion and vegetable seed’s USD 8.60 billion.
Gene editing is the only route to a royalty, and it is not in the forecast
The one technology that could change the structure is the first that could create defensible intellectual property in a crop where the public sector already owns the yield. The targets are well understood and commercially meaningful: semi-dwarf and lodging architecture, anaerobic germination, herbicide tolerance for direct seeding, blast susceptibility, cadmium accumulation and grain quality.
Japan operates a notification and consultation pathway and, on USDA Foreign Agricultural Service reporting from December 2025, nine genome-edited products have completed consultation, four of them in domestic production and distribution. Those are food and feed products across the category, not rice, and none has yet reset a market. Regulation (EU) 2026/1388, adopted on 17 June 2026, in force from 16 July 2026 and applying from 17 July 2028, creates a separate pathway for certain new genomic techniques in the world’s most restrictive market. It matters little to rice directly and a great deal as precedent.
The other option on the table is genetically modified rice, and it should be modelled as an option rather than a base case. China has held biosafety certificates on insect-resistant and herbicide-tolerant rice without commercial release for more than fifteen years while converting corn and soybean to routine commercial planting. Nothing in the current policy record suggests the decision arrives before 2033. If it did, it would be worth more to seed value per hectare than anything else in this report.
What would move the number
Substitution risk is high and does not fall through the window. Roughly two thirds of world rice area is sown with seed that never enters a commercial channel, a public variety is available at cost recovery in most producing countries, and any price increase a grower judges disproportionate to the yield difference is met with a reversion to saved seed within one season. The report assesses overall market risk as moderate, geopolitical exposure moderate and substitution risk high, and would move the overall assessment to high if Indian support prices were held flat for two consecutive seasons while world stocks stayed above 55 million tonnes.
The production side carries a risk that cannot be diversified by contract. Hybrid seed production is an inter-planted, flowering-synchronised operation that depends on wind, temperature and rainfall at a specific moment, concentrated in Hunan, Jiangsu, Jiangxi, Fujian, Sichuan and Hainan for China and in Telangana, Andhra Pradesh, Karnataka and eastern India for India. A typhoon or a heat event at flowering removes a season’s supply of a variety and cannot be substituted from elsewhere, because the multiplication windows are seasonal. In a market where germplasm is not defensible, multiplication capacity and the grower relationships that secure it are the only assets that are.
Six things to watch
- India’s hybrid penetration at the end of kharif 2026, and whether the 6 to 8 per cent acreage gain of 2025 becomes a trend or a procurement-price artefact.
- Whether Longping’s rice seed revenue turns positive in 2026, or whether the company completes its conversion into a corn business with a rice heritage.
- World milled production in 2026/27 at 533.08 million tonnes, and whether a tighter paddy price restores grower spending power in 2027.
- Direct-seeding adoption in Vietnam, Thailand and the Indian east, and the seed-rate effect of mechanised precision drilling against broadcast.
- The first commercial gene-edited rice variety outside Japan’s consultation pathway, and whether it creates enforceable rights in a crop that has never had any.
- Sub-Saharan African purchased-seed penetration, where a single percentage point of area is worth more to this market than a year of Chinese price recovery.
Scope, method and what this release does not do
The 2025 base was triangulated from three directions rather than licensed from one study: rice seed as 8.10 per cent of the USD 78.40 billion global seed market; bottom-up from disclosed tier-1 and national-champion revenue, principally Longping’s RMB 1,974 million converted at an average 2025 rate of 7.18, plus Bayer, Syngenta and SeedWorks positions in India and Southeast Asia; and per hectare from seed cost applied to national rice area with documented hybrid and varietal seed rates by establishment method. The three converge in a USD 5.9 to 6.8 billion band and the midpoint is taken. No licensed third-party market study is used or referenced.
The volume series measures commercial, purchased seed only, not total seed sown, and is expressed in million tonnes. Rice seed rate per hectare varies roughly sixfold between transplanted hybrid and broadcast dry direct-seeded systems, which makes a total-sown tonnage an arithmetic exercise rather than an analytical one. Rice is mapped to Wikidata Q5090, verified against the API. Bayer is mapped to Q812131 and Syngenta Group to Q114634360.
This release flags, without resolving, five inconsistencies in the underlying report. The published forecast describes rice as one of the three largest crop markets in the industry, and on the cluster’s own base values it is fourth, behind corn, soybean and vegetable seed. A published FAQ describes corn’s 31.4 per cent of the seed market as sitting on a fraction of the land, which contradicts the report’s own companion article on rice and corn being sown on almost the same area. The establishment split compounds to 1.7 per cent above the market total. The 36/64 decomposition is stated with no method, where a log decomposition and an endpoint bridge give different and both defensible answers. And the report’s own introduction and its own executive summary assign the gap between volume and value growth to conversion, which the identity shows is price and mix. None of the five changes the direction of the forecast. Two of them change its arithmetic.
Further reading
Global Rice Seed Market Report, Size, Share and Forecast 2026–2033 — the full report, with the regional split, the segmentation by seed class and establishment method, the company profiles and the forecast model.
Global Seed Market Report, the USD 78.40 billion hub this market sits inside at 8.10 per cent; corn seed at USD 24.60 billion, which is the comparison that explains the rate; and hybrid seed, whose rice germplasm line reconciles to the USD 2.44 billion hybrid class here.
Three articles in this cluster carry the argument further: Hybrid Rice Is a Chinese Market in Managed Decline and That Is the Entire Forecast; Hybrid Conversion Is Unfinished and India Is the Only Country That Can Finish It; and No Commercial GM Rice Exists Anywhere, and That Is an Export Story.
About this release
This press release accompanies the Pheonix Research report Global Rice Seed Market Report, Size, Share and Forecast 2026–2033, published in the Agriculture & Agritech industry, Seeds segment. Figures are Pheonix Research estimates for the 2025 base year unless attributed to a named source; company figures are taken from tier-1 disclosures. All percentages are calculated from the underlying values and rounded at the point of presentation, so component shares may not sum exactly to 100.
About Pheonix Research
Pheonix Research publishes market intelligence built from primary disclosure, official statistics and a documented forecast model, with each figure traceable to its source. Our agriculture coverage spans the seed, crop protection, fertiliser and biologicals markets, and the reports are designed to be read against each other rather than in isolation, because the same hectare, the same grower and the same support price sit behind all of them.
