Skip to article content

Who buys the most, in absolute terms

The Department of Fertilizers reports urea requirement, availability and sales for every state and union territory. On the sales line — the closest official proxy for consumption — India moved 357.81 lakh tonnes in 2023-24, 387.92 in 2024-25 and 396.60 in 2025-26. Availability exceeded requirement in all three years.

Uttar Pradesh alone accounts for 80.37 lakh tonnes, a little over a fifth of the national total. Madhya Pradesh follows at 42.43, then Punjab at 31.39, Rajasthan at 28.05 and Maharashtra at 27.27. On that reading the story looks like a simple one about big agricultural states buying a lot of fertiliser.

Divide by land and the ranking scrambles

Gross cropped area counts a field twice if it carries two crops in a year, which makes it the right denominator here: urea is applied per crop, not per acre of land owned. India’s gross cropped area in 2023-24 was 217.9 million hectares. Divide the urea across it and the national average comes to about 182 kg of urea per hectare — roughly 84 kg of actual nitrogen, since urea is 46 per cent N.

Urea sales 2025-26 from the Department of Fertilizers. Cropped-area shares derived from Land Use Statistics 2023-24; see the caveat at the end.

Three states move sharply. Rajasthan is India’s fourth-largest urea buyer in tonnes and its second-largest cropped area — and lands at 101 kg a hectare, 45 per cent below the national rate. Madhya Pradesh is second in tonnes and first in cropped area, and comes in at 140. Maharashtra crops a ninth of India and uses 113. Meanwhile Punjab, with 3.8 per cent of the cropped area, takes 7.9 per cent of the urea.

It is a crop map

The states above the line share a cropping pattern rather than a geography. Punjab, Haryana and western Uttar Pradesh run irrigated wheat-paddy rotations; Bihar runs paddy-wheat-maize; coastal Andhra Pradesh and Telangana run paddy and maize. Every one of those is a cereal system, and cereals are where nitrogen goes. Five crops — rice, wheat, maize, sugarcane and cotton — account for roughly 70 per cent of India’s total nitrogen consumption.

The states below the line grow something else. Rajasthan’s cropped area is dominated by bajra, mustard, gram and other rainfed crops with modest nitrogen demand, much of it unirrigated. Madhya Pradesh has swung heavily into soybean and gram; Maharashtra into cotton, soybean and pulses. Pulses and other legumes fix atmospheric nitrogen through root-nodule bacteria and need comparatively little applied N — a soybean or gram hectare is simply not a urea customer the way a paddy hectare is. Kerala, at the bottom on 44 kg a hectare, is mostly perennial plantation crops on a small cropped area.

Irrigation is the multiplier underneath all of this. Nitrogen only pays for itself if the crop has water to use it, and the intensity ranking tracks irrigation closely: Punjab irrigates 93.5 per cent of its gross cropped area and Haryana over 90, against a national figure near 60.

Put the crops themselves on the same axis and the state map stops being mysterious. The chart below is the most detailed public crop-wise split India has — from the FAO’s country study, and old enough that the absolute levels have since risen, but the ordering is the structural point.

Sugarcane takes about 125 kg of nitrogen a hectare, wheat 100, cotton 90, paddy 82. Pigeon pea takes 21, pearl millet 22, groundnut 24 — a third or less. That is not a difference in farmer behaviour; it is a difference in what the plant does. Pulses and groundnut host rhizobia in their root nodules that fix nitrogen from the air, so applied N is a starter dose rather than the main supply.

The right-hand column explains the state map directly. Paddy and wheat alone take 52.8 per cent of all the fertiliser India applies, on 70.4 million hectares. Add cotton and sugarcane and it is nearly two-thirds. A state’s urea intensity is very close to a weighted average of these crop numbers — which is why Punjab and Haryana, running wheat on paddy, sit where they do, and why Rajasthan’s bajra-and-gram acreage keeps it at 101 kg of urea a hectare despite being the second-largest cropped area in the country.

What ICAR actually recommends

Against those application rates it is worth setting the published Recommended Dose of Fertilisers. ICAR’s rice institute at Cuttack — ICAR-National Rice Research Institute, whose crop-advisory FAQ carries the doses in full — publishes these figures for rice, as N-P-K in kilograms per hectare. The urea column converts the nitrogen half at 46 per cent N.

Read the two tables together and the gap is visible. A hectare of irrigated kharif paddy is meant to receive about 130 kg of urea; a hectare of rabi rice about 174. Punjab is applying 379 kg per cropped hectare and Haryana 333 — and because gross cropped area already counts the second crop separately, those are per-crop-season figures, not annual totals across a double crop.

Two honest qualifications before that comparison is pushed too far. Urea is not the only nitrogen a field receives — DAP and NPK complexes carry N as well, which pushes actual nitrogen higher still rather than lower. And a state-average kg-per-hectare mixes crops with very different doses, so it cannot indict any individual farmer. What it can show is a systematic regional pattern, and it does.

The 4:2:1 norm is a north-west artefact

Indian fertiliser policy has been argued for decades against a target NPK ratio of 4:2:1, and the drift away from it is stark. The Economic Survey 2025-26 tracks the applied ratio from 4:3.2:1 in 2009-10, close to the benchmark, to 7:2.8:1 by 2019-20 and about 10.9:4.1:1 by 2023-24 — a deterioration it attributes overwhelmingly to excess nitrogen. A 2015 study by Ramesh Chand and Pavithra S — Chand then at NITI Aayog, both at ICAR’s National Institute of Agricultural Economics and Policy Research — took that norm apart, and the finding has held up.

They estimated, for every state, the nutrient quantity implied by multiplying the area under each crop by the dose recommended for that crop by the state agricultural universities and ICAR institutes. In other words they built a normative fertiliser requirement out of the actual cropping pattern — exactly the exercise the map above gestures at.

Two results matter. First, the national norm implied by India’s real cropping pattern is not 4:2:1 but about 2.6:1.4:1 — nitrogen 52 per cent of total nutrients rather than the 57.8 per cent the old norm implies. Second, and more pointedly: the 4:2:1 ratio came close to the estimated norm only in the Green Revolution belt of north-west India, because that is where the underlying agronomic trials were run, on the wheat and paddy that dominate its cropping pattern. Everywhere except Rajasthan, the crop-derived optimum implies less nitrogen and more phosphorus and potash than the national norm assumes.

Their state findings line up with the intensity ranking above to a degree worth noting, given the datasets are a decade apart. They found actual nitrogen use running above the crop-derived requirement in Andhra Pradesh, Punjab, Bihar, Haryana, Jharkhand and Assam — five of those six sit in the top seven of the per-hectare table. In twelve other states nitrogen use was below the recommended level, by anything from 4.4 to 60 per cent. India’s real deficits, on their estimate, were phosphorus and especially potash: against actual use of 3.26 million tonnes of K, the crop-derived requirement was 6.6 million.

That reframes the problem. The familiar claim that India over-applies urea is true in about a third of its states and false in the rest, and a single national ratio cannot express it.

What ICAR proposes instead of a blanket dose

The practices ICAR now issues are built around not applying a uniform dose at all. A December 2025 paper from ICAR-Indian Institute of Soil Science, Bhopal, sets out the current framework — “Smart nutrient management practices for sustainable crop production” in ICAR’s journal Indian Farming. It is the single most useful public statement of what the recommended practice actually is, and it is short.

Its starting point is efficiency, not quantity. Fertilisers deliver an estimated 40 to 50 per cent of India’s yield gains, but recovery is poor: 30 to 50 per cent for nitrogen, 15 to 20 per cent for phosphorus, 50 to 80 per cent for potassium and under 5 per cent for micronutrients. Half the urea does not reach the plant. Around that the paper sets the 4R principle — right source, right rate, right time, right place — and three delivery methods: Soil-Test Crop Response, which sets the dose from a soil test and a yield target rather than a blanket figure; Integrated Plant Nutrient Supply, combining mineral fertiliser with organic manures, crop residues and biofertilisers; and Site-Specific Nutrient Management. In-season decisions are meant to be guided by cheap instruments — leaf colour charts, SPAD meters, GreenSeeker sensors — that tell a farmer whether the crop needs the next split of nitrogen at all.

ICAR also publishes crop-level guidance on the same logic through its institutes: its integrated nutrient management guidance covers green manuring with sesbania, azolla in paddy, split application of nitrogen, and residue recycling in perennial crops. The operational delivery route for soil-test-based doses is the government’s Soil Health Card portal.

The distance between that framework and a national average of 182 kg of urea a hectare — 379 in Punjab — is the actual policy problem. It is not that the recommended doses are unknown. It is that a nitrogen fertiliser priced far below phosphorus and potash gives a farmer little reason to follow them.

The new organic category, and what GOBARdhan actually makes

The most concrete recent move on the other side of this ledger is regulatory rather than agronomic, and it came in two steps. On 17 February 2025 the Ministry of Agriculture and Farmers Welfare amended the Fertiliser (Control) Order, 1985 to create a new Schedule VIII category, “organic carbon enhancers from CBG plants”. The category existed but was empty. On 27 March 2025 a second notification — the Fertiliser (Inorganic, Organic or Mixed) (Control) (Third) Amendment Order, 2025 — filled it in, setting the actual specifications for Fermented Organic Manure and its liquid form.

The material in question is the by-product of compressed-biogas production. Every CBG plant digesting cattle dung, press mud or paddy straw produces bio-slurry alongside the gas: solid, it is FOM; liquid, LFOM. Until these notifications they sat awkwardly under the older biomanure and biofertiliser entries of the 2020 and 2021 gazette notifications, where they routinely failed the specification outright — the detail of which is in the next section. Product that cannot meet a specification cannot legally be sold as fertiliser, which is how CBG plants ended up sitting on it.

The volumes are not marginal. Estimates of how much of a CBG plant’s feedstock leaves as FOM or LFOM vary widely and are worth quoting rather than averaging: reporting on the amendment puts it at 88 to 92 per cent of total feedstock processed, while other industry descriptions cite 10 to 15 per cent. The gap is almost certainly a wet-versus-dry distinction — nearly everything that goes in comes out as wet digestate, of which only a fraction is saleable solid manure — but no single official figure resolves it. Either way, the disposal problem is concentrated in Punjab and Haryana, where straw-fed CBG capacity is highest and spare land to dump slurry on is scarce. The states at the top of the urea table are also the ones with the most unsold organic manure.

The money moves the other way now. Under the Department of Fertilizers’ Market Development Assistance scheme, ₹1,500 per tonne is paid on sales of FOM, LFOM and phosphate-rich organic manure produced at GOBARdhan biogas and CBG plants.

What the notification says a bag must contain

The specifications are worth setting out in full, because “organic manure” is otherwise an unfalsifiable claim. The figures below are from the gazette text itself — S.O. 1477(E) of 27 March 2025, as published on the GOBARdhan portal — not from summaries of it.

The striking thing is what the schedule does not require. There is no minimum nitrogen, no minimum NPK, no C:N ratio and no particle-size test anywhere in it. Both products are defined purely on organic carbon, pH, conductivity, pathogens and heavy metals — a carbon-and-safety standard, not a nutrient standard. That is the correct framing for what digestate is, and it is the clearest possible official statement that these are not urea substitutes. Read the FOM row against urea’s 46 per cent nitrogen and the argument is over.

Two drafting quirks are worth noting because they will confuse anyone reading a summary. The organic-carbon minimum is written in the gazette itself as a range, “12-14%”, which is unusual for a floor and is why secondary accounts quote both numbers. And several summaries circulating online transpose chromium and copper; the notified ceilings are chromium 50 and copper 300 mg/kg, in that order.

Two label obligations ride along: the instruction that FOM “shall be added 15-20 days prior to sowing”, printed on bags and containers, and a requirement that any nutrient enrichment be declared on the label.

Against that, the older entries these products used to sit under are what the industry says they kept failing — a 5 per cent NPK minimum and a 4 mm sieve at 70 per cent moisture that fibrous residue-derived FOM could not pass. Seen next to the gazette, the amendment did not lower a nutrient bar so much as remove one that was never the right test. Whether that reads as sensible standard-setting or as accommodation depends on what a buyer thinks they are purchasing; the case for the former is that nobody should have been buying digestate for its NPK.

ICAR’s package of practices, crop by crop

The specification says what may be sold. What to actually do with it is a separate document, and it is the most useful thing on the GOBARdhan portal: ICAR’s updated package of practices for FOM and LFOM across ten crops, filed as an action-taken report to a Secretary-level review of GOBARdhan on 5 July 2024. It is three pages, and it is specific.

Read down the right-hand column and the earlier point stops being an assertion. In eight of the ten crops ICAR pairs FOM with the full recommended dose of fertiliser. Only maize takes a reduction, to 80 per cent of RDF, and only spinach and pea are specified as genuine part-substitutions. This is the official Indian answer to “can organic manure replace urea”, and it is mostly no.

The doses also put the tonnage question in perspective. Wheat wants 6 to 8 tonnes of FOM a hectare; sugarcane wants 10 to 11 across two applications. India sold about 16.7 lakh tonnes of FOM, LFOM and PROM combined in 2025-26 — enough, at the wheat rate, to treat roughly two lakh hectares out of a wheat area of some 32 million.

ICAR closes with three instructions that apply to every crop: do not apply FOM or LFOM as a foliar spray; use it alongside the recommended dose of fertiliser; and keep the pH of the FOM between 6.5 and 7.5.

Two of those collide with the gazette, and a farmer following one document will be out of step with the other. ICAR’s general note says FOM should go in 7 to 10 days before sowing; the label mandated by the FCO says 15 to 20 days. And ICAR’s own spinach entry says 15 to 20, so the report is not internally consistent either. On pH, ICAR wants 6.5 to 7.5 while the notified specification permits 6.0 to 8.4 — a legally saleable FOM can sit outside the range ICAR recommends applying. Neither gap is dangerous, but both are the kind of thing that makes an extension worker’s job harder than it needs to be, and neither appears to have been reconciled.

Who tests it, and what the testing finds

Quality control runs through a three-stage structure that the FCO amendment of 13 December 2024 formalised. A fertiliser inspector draws three samples: one goes to the dealer or manufacturer, one to a designated state laboratory for the first analysis, and one is retained by an authority appointed by the state government.

If that first analysis returns non-standard, the affected party has 15 days to apply, with a fee, for a second analysis. The retained sample and the party’s own sample then go to the National Test House coding centre at Ghaziabad, which routes them for referee testing to NTH laboratories at Ghaziabad, Kolkata, Mumbai, Chennai or Jaipur. That report is normally final. Only where the first and second analyses call a sample non-standard on different parameters does the state have to send the remaining sample, on its own motion, to the Central Fertiliser Quality Control and Training Institute at Faridabad for a third and conclusive analysis. CFQC&TI dates to 1971-72 and runs three regional fertiliser control laboratories, at Navi Mumbai, Chennai and Kalyani, set up in 1988-89.

The gazette also specifies how each parameter is measured, by cross-reference to Part D of Schedule IV of the FCO: pH at serial 1, moisture at 2, organic carbon at 5, cadmium, copper, chromium, lead, nickel and zinc at 10 by atomic absorption spectrophotometry, mercury at 11, arsenic at 12, the pathogenicity test at 13 and conductivity at 14. Liquid FOM has to be oven-dried at 105°C before analysis, because the FCO’s parameters are defined on a dry-weight basis. Sampling itself follows Part A of Schedule II. The point of listing this is that the category is not a labelling exercise: there is a defined analyte list, a defined method for each, and a laboratory chain to run them.

The third-stage results are arresting and need reading carefully. In 2025-26 CFQC&TI received 1,078 samples and found 863 of them non-standard against 215 standard; in 2026-27 up to June, 172 of 223 were non-standard. More than 4,400 samples have gone through re-analysis since 2025.

Those look like catastrophic failure rates, and they are not. Third-stage samples reach Faridabad only after two laboratories have already disagreed about which parameter a sample failed — the referral is triggered by the disagreement, not selected at random. The pool is filtered for products that are already in trouble, so an 80 per cent failure rate at that stage says nothing about the failure rate in the market. What it does say is that the appeal route is not a rubber stamp for manufacturers: most challenges are lost. For an organic-manure category whose specification is new and whose producers are mostly small CBG operators rather than established fertiliser firms, that testing chain is the thing standing between a genuine soil amendment and a bag of wet digestate.

How much is actually moving

Offtake has grown fast from a very low base. Sales of FOM, LFOM and phosphate-rich organic manure ran 56,058 tonnes in 2023-24, 3.36 lakh tonnes in 2024-25, and 16.72 lakh tonnes in 2025-26 up to 28 February — more than 20.65 lakh tonnes cumulatively since the scheme began. On the supply side, 1,163 biogas plants and 426 CBG plants have registered on the GOBARdhan portal, and 44 memoranda of understanding have been signed between fertiliser marketing companies and CBG operators to sell the material. That last number is the one that matters, because it is the answer to whether the co-marketing route is real: it is, at 44 agreements.

Then set it beside the urea column. India sold 396.6 lakh tonnes of urea in 2025-26 and about 16.7 lakh tonnes of all three organic manures combined — roughly four per cent of the tonnage, and on nutrient content, a rounding error. The growth rate is genuinely steep; the base is genuinely tiny.

One figure does not reconcile, and is worth flagging rather than smoothing. Total Market Development Assistance released is ₹111.72 crore — ₹28 crore in 2024-25 and ₹83.72 crore in 2025-26 to 28 February. At ₹1,500 a tonne that covers about 7.4 lakh tonnes, against 20.65 lakh tonnes of reported sales. Either disbursement lags the claims badly, or a majority of the volume sold is not drawing the assistance at all. The published figures do not say which, and the difference matters: one is an administrative backlog, the other means the subsidy is not reaching most of the product it was designed to move.

Blend it, do not swap it

The practice guidance attached to all of this is narrower than the enthusiasm around it, and the distinction matters. The claim ICAR’s work supports is that biofertilisers raise crop yields by 10 to 25 per cent and can substitute for roughly 20 to 25 per cent of chemical nitrogen and phosphorus — and the operative words are used along with. This is a supplement to a reduced mineral dose, not a replacement for it. Nothing in that finding supports dropping urea and applying FOM instead; a hectare of irrigated paddy needing 60 kg of N does not get it from manure at any realistic application rate.

The one hard, official application instruction is on the label. The February 2025 amendment requires every manufacturer to state on the container that FOM and LFOM must be applied 15 to 20 days before sowing. That timing is the whole practice in a sentence: fermented manure has to mineralise in the soil before the crop can draw on it, so applied at sowing it does nothing for the young plant, and applied instead of a basal dose it leaves the crop short at exactly the wrong moment. Split the mineral nitrogen as usual; put the organic fraction in ahead of the season.

Two caveats worth carrying. The notification created the category but, as commentary at the time noted, did not itself settle revised quality standards for organic carbon enhancers — and without a standard, a farmer buying a bag has no way to know what is in it. And the commercial question is still open: whether the established fertiliser companies will actually procure and co-market FOM alongside their urea and DAP is the difference between a scheme and a supply chain. The blending advice is sound; the blend has to reach a field for it to matter.

Set against the numbers earlier in this piece, the arithmetic is sobering rather than encouraging. Substituting a quarter of the nitrogen on Punjab’s 379 kg of urea a hectare still leaves it near 285 — well above the recommended dose for either crop in its rotation. Biofertiliser blending narrows the gap between actual and recommended nitrogen. It does not close it, because the gap was never mainly an organic-matter problem. It is a price problem.

What a bag actually costs

All of the agronomy above runs into a price sheet. A 45 kg bag of urea carries a statutorily notified maximum retail price of ₹242, exclusive of neem-coating charges and taxes, and that number has not moved since March 2018. It works out to about ₹5.38 a kilo, or roughly ₹5,378 a tonne. The bag itself was cut from 50 kg to 45, and in some cases to 40, expressly to curb over-application — a nudge on quantity rather than on price.

What the bag costs to make or import is another matter: roughly ₹2,200, against the ₹242 the farmer hands over. The government carries 85 to 90 per cent of it, about ₹31,000 to ₹35,000 a tonne, and budgeted ₹1.19 lakh crore for urea subsidy in 2025-26 alone.

Set side by side, the two columns look deceptively close: ₹5.38 a kilo of urea against the ₹5.50 the Indian Biogas Association has argued FOM needs to be viable. They are not close at all. Urea is 46 per cent nitrogen, so its nitrogen costs the farmer about ₹11.70 a kilo. FOM has to clear only 5 per cent total NPK, so on any nutrient-for-nutrient basis it is an order of magnitude more expensive.

That comparison is unfair to FOM on its own terms — its value is organic carbon and soil structure, not NPK, and nobody buys manure as a nitrogen source. But it is the comparison a farmer standing in a shop makes implicitly, and it explains the adoption problem better than any awareness campaign will. One product is priced at a tenth of what it costs to supply; the other is priced at roughly what it costs to supply. A ₹1,500-a-tonne assistance does not bridge that, because the gap is not on FOM’s side of the ledger.

Raise the price, refund the farmer

The government’s own economists have now said the quiet part. The Economic Survey 2025-26, published in January 2026, recommended raising the retail price of urea outright — noting it has not moved from ₹242 a bag since 2018 — and pairing the increase with an equivalent direct benefit transfer to farmers on a per-acre basis.

The logic is worth stating precisely, because it is not an austerity argument. In the Survey’s own framing, farmers “receive the same overall purchasing power, but the relative price of nitrogen moves closer to its agronomic cost”. The money is not withdrawn; it is detached from the act of buying nitrogen. A per-acre transfer pays the same whether a farmer applies 130 kg of urea a hectare or 379, so the incentive to apply the 380th kilo disappears while the farm’s income does not. That is the entire mechanism, and it is the only instrument in this piece that acts on the actual binding constraint — price — rather than on information, labelling or exhortation.

It would also change the arithmetic for everything discussed above. Soil-test-based dosing, the 4R framework, FOM blending and the ₹1,500-a-tonne assistance all ask a farmer to use less of something priced at a tenth of its cost. Move the relative price and they stop being appeals to good practice and start being cheaper than the alternative.

The Survey flags one design problem, and it is a serious one: tenancy. A per-acre transfer tends to land with the landowner on record, not the tenant who actually buys and applies the fertiliser. In states where informal tenancy is widespread and unrecorded, that is not a detail to be worked out later — it decides whether the compensation reaches the person facing the higher price at the counter.

One more tension is worth naming, inside the government’s own position. The Survey measures the imbalance against the 4:2:1 benchmark, while the NITI-hosted research discussed earlier concludes that this benchmark holds only for the north-west and that the crop-derived national norm is nearer 2.6:1.4:1. Both point the same way on urea pricing. They disagree about what balance would look like once the price moved — and on the second view, a uniform national correction would still be wrong in most states, just less wrong.

What this piece could not establish

The per-hectare column is a derived figure, not an official one. The Department of Fertilizers publishes urea sales by state; the Directorate of Economics and Statistics publishes cropped area by state. Nobody publishes the ratio, so it is computed here — and the state cropped-area shares were read from the Land Use Statistics chart, which reports them to one decimal place. For the large states that rounding is immaterial; for Jharkhand and Kerala, on shares under 1.5 per cent, treat the figure as indicative. The urea year (2025-26) and the cropped-area year (2023-24) also do not match, because 2025-26 land-use data does not exist yet.

On the price side, one thing is absent rather than uncertain: there is no government-recommended price for FOM or LFOM published on the GOBARdhan portal. That site — gobardhan.co.in, run by the Department of Drinking Water and Sanitation — is a registration system for biogas and CBG plants, sits behind a login, and was under scheduled maintenance when this piece was written. What the government sets is the ₹1,500-a-tonne Market Development Assistance, which is a payment to the producer, not a notified price to the farmer. Unlike urea, FOM has no statutory MRP at all; the ₹0.50 to ₹4.50 range is observed market pricing and the ₹5.50 figure is an industry ask, not policy. Anyone told there is an official FOM price should ask to see the notification.

The specification and package-of-practices figures in this piece are taken from the gazette notifications and the ICAR report as published on the GOBARdhan portal, not from summaries, and both are linked above so a reader can check them. The ₹111.72 crore MDA figure against 20.65 lakh tonnes of sales remains unresolved: the gap is arithmetic from two published numbers, and this piece could not determine whether it reflects a disbursement lag or volume that never claimed the assistance.

Two further gaps. Sales are not consumption — stock moves across state lines and some urea is diverted to non-agricultural use, which the Chand and Pavithra paper explicitly flags as inflating apparent surpluses. And there is no current public crop-wise split of nitrogen use by state; the crop attribution here rests on each state’s cropping pattern and the published national figure that five crops take about 70 per cent of nitrogen, not on measured crop-level application data.